Anti-listerial peptides from a marine Bacillus velezensis FTL7: production optimization, characterizations and molecular docking studies

Abstract

Antimicrobial peptides (AMPs) with potent anti-listerial activity were characterized from a novel marine Bacillus velezensis FTL7. A Box-Behnken statistical experimental design was used to study the combined impact of culture conditions on the production of AMPs by B. velezensis FTL7. The conditions optimized by statistical experimental design were 34.5 °C incubation temperature, 23 h incubation time, and 7.6 initial pH of the medium. AMP purification was performed by ammonium sulphate fractionation and butanol extraction followed by reversed-phase C₁₈ solid-phase extraction. Tricine-SDS-PAGE analysis revealed a peptide with a molecular mass of ~ 6.5 kDa in an active AMPs fraction, whereas the mass spectrometry (MS) analysis showed the presence of AMPs in the mass range of 1–1.6 kDa, along with a 6.5 kDa peptide. Both MS and MS/MS analysis confirmed the AMPs as lipopeptides including surfactin, fengycins and iturin A and a circular bacteriocin amylocyclicin. The minimum inhibitory concentration of these AMPs against L. monocytogenes Scott A was 2.5 µg/mL. Further, the in-silico docking studies showed that the AMPs from B. velezensis FTL7 have high binding energy and stable binding patterns towards L. monocytogenes target proteins. Thus, this new combination of AMPs can serve as an effective food bio-preservative.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13205-024-03944-5.

Introduction

Listeria monocytogenes is a highly harmful foodborne pathogen that can proliferate in a wide range of substrates and is exceptionally resistant to various ecological stresses. It is ubiquitously present in nature and through contaminated food; it consumes by human beings and can give rise to life-threatening illnesses (Kujik et al. 2011; Carvalho et al. 2015; Wu et al. 2022). Intake of food contaminated with L. monocytogenes causes an acute disease known as listeriosis. Infection of L. monocytogenes leads to abortion or postnatal health complications, especially in immunocompromised individuals such as newborns, the elderly and pregnant women (Carvalho et al. 2015; Wu et al. 2022; Shen et al. 2023). The bacterial proteins internalin A and B help to induce their absorption into non-phagocytic cells by unique attachment to host exterior receptors (Pizarro-Cerdá et al. 2012). Also, L. monocytogenes can withstand extreme conditions such as salt (NaCl) concentration up to 25.5% and can grow over a wide range of temperatures (4–50 °C) and pH (4.3–9.8) (Kuijk et al. 2011; Wu et al. 2022). The most common methods for controlling L. monocytogenes in the food industry are heat treatment, irradiation, ozone, pulsed ultraviolet and chemicals (chlorine, hydrogen peroxide and thiamine dilauryl sulphate). However, these systems can have undesired effects, which are against food industries and consumers’ demand who ask for additive-free, fresher and more natural taste food products. The addition of chemical preservatives in food processing might affect human health and increase the risk of cancer (Wu et al. 2022). Therefore, the primary concern in the food industry is the development of safe, efficient and natural approaches to controlling the L. monocytogenes in foods.

Various naturally produced antimicrobial peptides (AMPs) have been studied as promising methods for averting the development of L. monocytogenes in food products that could improve food safety. It benefits producers and consumers in the food industry (Martín et al. 2022; Shen et al. 2023; Xia et al. 2023). All multicellular organisms synthesize various classes of molecules/compounds such as AMPs as part of their primary defence mechanisms. These peptides exhibit a wide range of antibacterial, antifungal, antiviral and anticancer activities (Maria-Neto et al. 2015; Johny and Suresh 2022). These are ribosomally synthesized peptides, even though some are extensively modified post-translationally (Chikindas et al. 2018) inside the cell or exterior during transport to form their biologically active conformation (Lettieri and Brandelli 2013). A large diversity of AMPs was reported from Bacillus sp., including bacteriocins and lipopeptides (LPs) (Ahire et al. 2020; Sakthivel et al. 2023). Bacteriocins are grouped into four main classes based on their thermal stability, molecular weight (size), chemical structure/moiety, etc. (Rungsirivanich et al. 2021): Class 1–lanthionine containing and thermostable peptides with < 5 kDa size, Class II–non-lanthionine thermostable peptides with < 10 kDa size, Class III–thermolabile peptides with > 30 kDa size and Class IV–complex peptides with a single lipid or carbohydrate moiety. LPs are grouped into three major families named (i) iturins, (ii) surfactins and (iii) fengycins (Ongena and Jacques 2008; Xu et al. 2018). The LPs have a cyclic configuration of seven to ten amino acids linked to a fatty acid derivative. Each LP type possesses a variety of isomers with diverse lengths of fatty acid chains or different combinations of amino acids (Fira et al. 2018; Liu et al. 2020; Johny and Suresh 2022). The amphiphilic character of the LPs, along with their synergy with the target cell membrane, accounts for their mode of action, which disturbs the configuration and permeability of the cell membrane of target bacteria (Fira et al. 2018). To combat different microbes in nature, most of the LP-synthesizing Bacillus sp. can produce more than one family of LPs (Liu et al. 2020).

Given the broad spectrum of activity and general biosafety, AMPs are emerging as potential substitutes for synthetic food preservatives. Nisin is a polycyclic AMP (class I bacteriocin) produced by a few stains of Lactococcus lactis subsp. Lactis, with a broad spectrum of antibacterial activity, and is widely used worldwide as a food preservative (Wei et al. 2021). It has prompted the interest of researchers in searching for novel/new AMPs with strong inhibitory activity towards different foodborne pathogens. Even though several AMPs have been isolated and characterized from various species of bacteria (Qin et al. 2019; Wei et al. 2021), their commercial exploitation for producing AMP is currently restricted due to the low yield and their action and stability. Marine bacteria have been documented as excellent sources of AMP, and they have been associated with various roles in defence and survival (Johny and Suresh 2022). However, as AMP synthesizers, marine bacteria have yet to be explored as much as their terrestrial counterparts. The data available on the process factors that affect AMP synthesis by marine bacteria under any bioprocesses are limited. Improvement of bacterial AMP production has gained much attention in current years owing to its promising commercial demand. Response surface methodology (RSM) is a collection of potent statistical design methods for evaluating various process factors concurrently since minimal experimental runs are required as related to the ‘one-factor-at-a-time’ methodology (Thadathil et al. 2014; Yue et al. 2021). The statistical design method of RSM has been widely applied in different stages of process optimization in microbial bioprocessing (Thadathil et al. 2014; Suganthi and Mohansrinivasan 2015; Yue et al. 2021; Emon et al. 2023). In a previous study, we isolated a Bacillus sp. (Bacillus velezensis FTL7) with a broad spectrum of antibacterial activity against various foodborne pathogens from a marine source (Johny and Suresh 2022). The complete genome sequence (GenBank accession number: CP094618) of B. velezensis FTL7 was examined with the help of BAGEL-4 and anti-SMASH to search potential AMP-synthesizing gene clusters, which revealed the presence of uberolysin/carnocyclin family circular bacteriocin-encoding gene and LP synthetase genes (Johny and Suresh 2022). In this study, the process factors/variables for the enhanced production of AMPs by B. velezensis FTL7 were optimized using a statistical experimental design. Further, the AMPs from B. velezensis FTL7 were purified, characterized and explored the anti-listerial activity. In addition, in-silico studies were performed to elucidate the probable binding mode and effectiveness against the selected listerial target proteins.

Materials and methods

Culture media and analytical reagents

Nutrient broth (NB); nutrient agar (NA); Zobell marine broth (ZB) and Zobell marine agar (ZA) (Hi-media, Mumbai, India); sodium chloride; ammonium sulphate; sodium dodecyl sulphate (SDS); trifluoroacetic acid (TFA); resazurin, glutaraldehyde and butanol (SRL, Mumbai, India); MS-grade water and acetonitrile (Honeywell Research Chemicals, Seelze, NI, Germany); ultra-low range-molecular-weight (Mw) protein marker, tricine and formic acid (Sigma Aldrich Co., St. Louis, MO, USA) and Sep-Pak C₁₈ cartridges (Waters, Milford, MA, USA) were used. All other chemicals and reagents were of analytical grade.

Bacterial cultures and maintenance

The marine bacterium B. velezensis FTL7 (GenBank accession number: CP094618) used in this study was previously isolated and identified (Johny and Suresh 2022). Prof. AK Bhunia, Purdue University, USA, gifted the indicator bacteria L. monocytogenes Scott A. B. velezensis FTL7 and L. monocytogenes Scott A were maintained in ZA/ZB media and NA/NB media, respectively. The stock of both cultures was stored in 40% glycerol at − 80 ± 1 °C. For analysis, aliquots of FTL7 in ZB at 32 ± 2 °C and L. monocytogenes in NB at 37 ± 2 °C were grown under constant shaking conditions (115 rpm).

Production of AMPs from B. velezensis FTL7 and optimization of process parameters by design experiment

Production of AMPs from B. velezensis FTL7 under the shaken flask cultivation method was carried out following the procedure reported previously (Johny and Suresh 2022). Briefly, 100 mL of ZB in 250 mL Erlenmeyer conical flask was inoculated with 100 µL of inoculum and incubated on a shaking incubator at 120 rpm for 24 h at 32 ± 2 °C. After incubation, the cell-free supernatant (CFS) was collected and tested against L. monocytogenes Scott A by disc diffusion assay (Johny and Suresh 2022) and stored at − 20 °C for further analysis.

Selection of variables

Selecting a suitable level of experimental parameters/variables helps to enhance the yield of AMP from B. velezensis FTL7. Variables and their levels were chosen based on the results of preliminary experiments. In the initial investigation, incubation temperature (20–45 °C), pH of the culture medium (5.6–9.6) and incubation time (6–38 h) were applied and tested using the one-variable-at-a-time experiments.

Box-Behnken design

Box-Behnken design experiments with three-factor-three level (− 1, 0 and + 1) were carried out to analyse the effect of different independent variables and their interactions on the yield of AMP from B. velezensis FTL7 on the basis of the results of preliminary one-variable-at-a-time experiments. The design model was applied to the independent variables/factors, including incubation temperature (X1; °C), incubation time (X2; h) and pH of the culture medium (X3) with one block and 15 experimental runs (Table 1). The yield of AMP measured as the zone of inhibition (in mm) against L. monocytogenes Scott A was considered as the dependent/response variable (Y1). Growth of B. velezensis FTL7 measured as the optical density of culture broth at 600 nm (OD @ 600 nm) was regarded as a second dependent/response variable (Y2). The linear association among the independent variables and the response variable was fitted into the 2-order polynomial equation (Eq. 1) to achieve the regression coefficients. For all design experiments, the Erlenmeyer conical flask (100 mL) with 40 mL of medium with 0.25% (v/v) inoculum (1 OD @ 600 nm = 1 × 10⁸ CFU/mL) of B. velezensis FTL7 was used. To verify the aptness of the model, the production of AMP at the optimized conditions was carried out, and the results were compared with the predicted value determined by the regression equation. The Statistica version 7.1 statistical software package was used to design the experiments and analyse the data (Statsoft 2005).

Table 1.

Independent variables in coded and actual levels (A) and Box-Behnken design matrixes with independent variables at coded and actual values and response variable of antibacterial activity (zone of inhibition, mm) at observed and predicted values (B)

A
Symbol	Independent variables		Level
	Variables	Unit	− 1	0	1
X1	Temperature	°C	27	32	37
X2	Time	h	01	12	23
X3	pH	–	6.6	7.6	8.6

B
Run	Independent variables						Response variable
	X1		X2		X3		(antibacterial activity, mm)
	Coded	Actual	Coded	Actual	Coded	Actual	Observed	Predicted
1	(− 1)	27	(− 1)	1	(0)	7.6	0.00 ± 0.00	0.15
2	(1)	37	(− 1)	1	(0)	7.6	0.00 ± 0.00	2.74
3	(− 1)	27	(1)	23	(0)	7.6	10.25 ± 0.24	7.51
4	(1)	37	(1)	23	(0)	7.6	9.90 ± 0.15	9.75
5	(− 1)	27	(0)	12	(− 1)	6.6	0.00 ± 0.00	1.12
6	(1)	37	(0)	12	(− 1)	6.6	10.00 ± 0.04	8.53
7	(− 1)	27	(0)	12	(1)	8.6	0.00 ± 0.00	1.47
8	(1)	37	(0)	12	(1)	8.6	0.00 ± 0.00	− 1.12
9	(0)	32	(− 1)	1	(− 1)	6.6	0.00 ± 0.00	− 1.27
10	(0)	32	(1)	23	(− 1)	6.6	8.60 ± 0.07	10.22
11	(0)	32	(− 1)	1	(1)	8.6	0.00 ± 0.00	− 1.62
12	(0)	32	(1)	23	(1)	8.6	0.00 ± 0.00	1.27
13	(0)	32	(0)	12	(0)	7.6	10.50 ± 0.03	10.40
14	(0)	32	(0)	12	(0)	7.6	10.30 ± 0.04	10.40
15	(0)	32	(0)	12	(0)	7.6	10.40 ± 0.04	10.40

Purification of AMPs from B. velezensis FTL7

All experiments to purify the AMPs from B. velezensis FTL7 were carried out at 4 °C using a walk-in cold room. AMPs from the CFS of FTL7 were initially concentrated by ammonium sulphate fractionation (20–80% saturation), followed by butanol extraction (Sharma et al. 2021). The resultant precipitate from each ammonium sulphate fractionation was collected by centrifugation (12,000 rpm, 10 min, 4 °C) and re-suspended in a minimum quantity of sodium phosphate buffer (100 mM, pH 7.0). The antibacterial activity of each fraction was tested by disc diffusion assay (Johny and Suresh 2022). Further, the active fraction thus obtained was added with n-butanol (1:1, v/v) and set aside at 4 °C under mild continuous mixing using a magnetic stirrer for 1 h and centrifuged (8000 rpm, 10 min, 4 °C). Then, the butanol upper portion was collected, lyophilized (sample 1) and stored at − 20 °C for further analysis. The antibacterial activity was confirmed by disc diffusion assay (Johny and Suresh 2022).

The purity of sample 1 was tested by Tricine-SDS-PAGE analysis as described by Schägger and Von Jagow (1987). Briefly, 40 µg of sample 1 was mixed with sample loading buffer (1:3, v/v) and incubated for 15 min at 75 ± 2 °C, Further, the sample mixture was loaded onto Tricine-SDS-PAGE gel (4% separating and 16% resolving gel) with 50 µL per well. One part of the gel was stained with Coomassie brilliant blue R-250, and another part of the gel was used for antibacterial activity by agar-overlay methods. The agar-overlay assay of AMP-containing Tricine-SDS-PAGE gel was carried out as per the procedure described by Yu et al (2012). Briefly, the electrophoresed gel was fixed with isopropanol, acetic acid and water (20:10:70, v/v/v) for 45 min and washed with sterile distilled water for 2 to 3 h. The processed gel was overlayed with 0.5% NA medium containing the cells of L. monocytogenes (CFU/mL ~ 10⁸) and incubated at 37 ± 2 °C for 12–18 h. After incubation, the prepared gel was checked for the antibacterial activity of the peptide band.

The freeze-dried AMP sample (sample 1) was dissolved in a minimum quantity of sodium phosphate buffer (100 mM, pH 7.0) and subjected to further purification by reversed-phase chromatography on Sep-Pak C₁₈ cartridge (500 mg). The cartridge was equilibrated with 100% acetonitrile (100%) followed by ultrapure (Type 1) water containing 0.1% (v/v) TFA. Further, the enriched AMP sample (sample 1) was loaded onto the equilibrated cartridge. After washing with ultrapure (Type 1) water containing 0.1% (v/v) TFA, elution was carried out at 50% and 80% (v/v) acetonitrile. The eluates were concentrated using a vacuum concentrator, suspended in a minimum volume of ultrapure (Type 1) water and stored at − 20 °C. All the eluates were tested for antibacterial activity by disc diffusion assay (Johny and Suresh 2022). The fraction which showed the antibacterial activity was named sample 2.

Mass spectrometry (MS) analysis of peptides

The active fraction of AMPs (sample 2) was dried in-vacuo and suspended in 50% acetonitrile containing 0.1% (v/v) formic acid. The sample was directly injected into a Triple TOF 5600 + mass spectrometer (Sciex LP, ON, Canada) operated in positive ion mode with the following MS parameters: curtain gas–25, temperature–400, ionspray voltage floating–5500 V, declustering potential–100 and collision energy–10. TOF MS range was 100–3000 m/z. For tandem  experiments (MS/MS), precursor ions were manually selected and subjected to a collision energy of 30–40. The mass spectra were evaluated manually using PeakView 2.1 software (Sciex LP, ON, Canada) (Epparti et al. (2022).

Minimum inhibitory concentration and time-kill assay

The minimum inhibitory concentration (MIC) of the AMPs from B. velezensis FTL7 was calculated using the microtiter plate dilution approaches (Wie et al. 2021). Briefly, the serially dispersed AMP sample to a total of 0.1 mL medium in a multiwell-plate was added with 0.1 mL of inoculum of L. monocytogenes Scott A (18 h aged in NB, 0.5 OD @ 600 nm) and incubated at 37 ± 2 °C. After 18 h of incubation, the growth was observed at OD₆₀₀ using a microplate reader and the MIC (µg/mL) of the AMP sample was determined (Kujik et al. 2011; Epparti et al. 2022). Further, to determine the viability of the test bacterium (L. monocytogenes Scott A), the culture broth in each well was mixed with 30 µL resazurin dye (0.015%, w/v) and the viability of the L. monocytogenes was observed (Teh et al. 2017).

Time-kill assay of AMPs from B. velezensis FTL7 was carried out against L. monocytogenes (Singh et al. 2021; Wei et al. 2021). The actively growing cells from a broth culture (18 h) of L. monocytogenes (CFU/mL ~ 10⁸) in NB were collected aseptically by centrifugation (8000 rpm, 10 min) and washed repeatedly using sterile phosphate-buffered saline (PBS). The washed cell pellet was re-suspended in sterile PBS and treated with 1 × MIC and 0.5 × MIC of purified AMPs and incubated at 37 ± 2 °C for 30–180 min. Aliquots (0.1 mL) from each treated sample were taken at different time intervals and the cells were collected by centrifugation (8000 rpm, 10 min). After serial dilution, the washed cell pellets were plated on pre-solidified NA. Untreated L. monocytogenes cells were used as the control. After incubation for 24 h at 37 ± 2 °C, the plates were observed for any growth of the indicator organism and the CFU was calculated for plotting the graph (Singh et al. 2021; Wei et al. 2021).

Scanning electron microscopy (SEM) analysis of L. monocytogenes Scott A by AMPs treatment

The L. monocytogenes Scott A was grown in NB for 18 h (37 ± 2 °C), and the cells were harvested by centrifugation (8000 rpm, 10 min) and washed with sterile physiological saline. The cell pellet was re-suspended in sterile physiological saline. The purified AMPs (10 µg/mL) from B. velezensis FTL7 were added to the cell suspension and incubated for 2 h at 37 ± 2 °C and centrifuged at 8000 rpm for 10 min. The collected cells after centrifugation were washed with sterile physiological saline and mixed with 2% glutaraldehyde. After overnight storage at 4 °C, the fixed cells were washed with sodium phosphate buffer (100 mM, pH 7.0). Further, the cells were dehydrated using a series of gradient ethanol (10–100%) for 15 min at 37 ± 2 °C and dried (Lee and Chang 2018). The cells were placed on a double-sided conductive tape fixed onto the sample holder and observed using the SEM (Carl Zeiss Microscopy, NY, USA) at 15 kV.

In-silico analysis

Protein-LP docking

The X-ray crystallography structure of L. monocytogenes surface protein internalin (PDB ID: 1O6S), listeriolysin (PDB ID: 4CDB) and transcriptional regulator protein, GntR (PDB ID: 4R1H), was retrieved from the Protein Data Bank. The LPs-surfactin (PubChem CID: 443592), fengycin (PubChem CID: 443591) and iturin A (PubChem CID: 102287549) were retrieved from the PubChem database as Structure Data File (SDF) (Fig. 5a). To assess the interactions of LPs with target proteins, automated docking was done by employing Autodock Vina software by the method described by Aditya Rao and Shetty (2022). Briefly, as a part of the target preparation, the Gasteiger charges, C-terminal oxygen and polar hydrogen were added to the protein structures. The residues composing the active pocket were enclosed by the grid plot (Online resource 1f). For surfactin, fengycin and iturin A, all the torsions were permitted to revolve during docking, and the atomic coordinates were stored along with their partial charges.

Fig. 5.

Three-dimensional structure of surfactin (A), fengycin (B) and iturin A (C) (a); Interaction of surfactin (A), fengycin (B) and iturin A (C) with internalin, interaction of surfactin (D), fengycin (E) and iturin A (F) with GntR, and interaction of surfactin (G), fengycin (H) and iturin A (I) with listeriolysin (b)

Peptide modelling and protein-bacteriocin docking

As there was no structure file available for the cyclic bacteriocin amylocyclicin, the peptide structure was modelled prior to interaction studies with targeted proteins (internalin, listeriolysin and GntR). For homology modelling, Phrye2, an online protein modelling server, was used, and ZDOCK version 3.0.2 was employed to perform the rigid-body docking (Chen et al. 2003). The top-ranked structure arranged based on an energy-based scoring system was selected out of the top 10 predictions of complex structures displayed as ZDOCK results.

Statistical analysis

All the studies and calculations were carried out in triplicate (except SEM and MS) and were reported average value ± standard deviation. The statistical analysis (one-way analysis of variance, ANOVA) was performed to analyse statistical significance. p ≤ 0.05 denoted the study results with statistical significance. The Statistica version 7.1 statistical software package was used in the investigation to analyse the experimental data (Statsoft 2005).

Results and discussion

AMP production by B. velezensis FTL7

B. velezensis FTL7 isolated from marine sediment could grow well under shaken flask culture conditions and produce AMP with broad antibacterial activity (Johny and Suresh 2022). The influence of different growth factors, such as incubation temperature, incubation time and initial pH of the medium, in the production of antimicrobial compounds by bacteria has been reported (Suganthi and Mohansrinivasan 2015; Lahiri et al. 2021). Hence, the present study aimed to analyse the effects of the initial pH of the medium, incubation temperature and incubation time on the growth of FTL7, and synthesis of AMPs was studied using one-variable-at-a-time experiment approaches (Fig. 1). Data presented in Fig. 1a show that the initial pH of the culture medium had a significant (p ≤ 0.05) effect on growth and AMP production by FTL7. Both growth of the culture and AMP production were observed in a medium with an initial pH of 7.6. The ideal incubation temperature for the enhanced production of AMP by FTL7 was 32 ± 2 °C (Fig. 1b). As shown in Fig. 1c, the AMP production and growth of FTL7 were observed maximum at 20–24 h of the incubation time. Production of AMP declined beyond 24 h of incubation. Data presented in Fig. 1 suggest that the production of AMP by B. velezensis FTL7 was growth associated.

Fig. 1.

Production of the antibacterial peptide by one-variable-at-a-time experiment (a–c) and response surface plot as a function of incubation time and incubation temperature (d), response surface plot as a function of medium pH and incubation temperature (e), response surface plot as a function of medium pH and incubation time (f) at the central value of other parameters on antibacterial peptide production by Bacillus velezensis FTL7

Optimization of process variables on AMP production using design experiments and model development approaches

In general, the yield of AMP is very low, and its synthesis by any microorganism is greatly influenced by multiple factors, and they are usually strain specific (Carolissen-Mackay et al. 1997). The combined effect of multiple factors, including growth-associated and stress-associated mechanisms that regulate the mechanism of AMP formation, results in the production of AMPs (Lahiri et al. 2021). These factors must be optimized along with the selection of a suitable strain to enhance its yield (Suganthi and Mohanasrinivasan 2015; Lahiri et al. 2021; Yue et al. 2021).

The main factors of initial pH of the medium, incubation temperature and incubation time had a significant impact on the production of AMPs by B. velezensis FTL7. Therefore, for the optimization of AMPs, the effects of these three main were further studied using a statistical experimental design. The predicted values of response variable Y1 (AMP yield), along with the observed values of the design experiment, are presented in Table 1. ANOVA was used to analyse the response (Y1) obtained from the experiment (Table 1). Significant factors were defined by values p ≤ 0.05. The result of ANOVA of the experimental data of response variable Y1 is presented in Online resource 1. The predicted values of response variable Y2 (growth of B. velezensis FTL7), along with the observed values of the design experiment and ANOVA, are presented in Online resource 2. Data presented in Table 1 and Online resource 1 suggest that the regression model used to clarify the association of AMP production (Y1) with the variables evaluated was precise. The effectiveness of fit of the model equation was examined by the determination of coefficient (R²) (0.91919), which depicts that the response model can clarify a total variation of 91.9% for AMP production (Y1) and also expressed that the design can effectively be applied and accepted. The 2-order polynomial equation derived from the design experiment is presented in Eq. 1.

$$\text{Y1}=\left(-541.205\right)+\left(10.476\ast \text{X1}\right)+\left(-0.100\ast {\text{X1}}^2\right)+\left(2.430\ast \text{X2}\right)+\left(-0.024\ast {\text{X2}}^2\right)+\left(98.005\ast \text{X3}\right)+\left(-5.394\ast {\text{X3}}^2\right)+\left(0.002\ast \text{X1}\ast \text{X2}\right)+\left(-0.500\ast \text{X1}\ast \text{X3}\right)+\left(-0.195\ast \text{X2}\ast \text{X3}\right)$$ 1

where Y1 stands for the response value (yield of AMP, measured as the zone of inhibition against L. monocytogenes Scott A), while X1, X2 and X3 denote incubation temperature, incubation time and pH of the medium, respectively.

ANOVA (Online resource 1a) and the Pareto chart (Online resource 1b) for the three independent factors explained that the quadratic model resulting from the Box-Behnken design could sufficiently be applied to illustrate the factors for AMP production (Y1) under a broad choice of the experimental conditions. Of the three independent variables considered, incubation temperature had a quadratic (p < 0.009) and pH of the culture medium had a linear (p < 0.008) effect on AMP production (Y1) by B. velezensis FTL7 (Online resource 1a). Figure 1d–f are the response surface and counterplots of AMP production (Y1), which clarifies the association between response and the experimental data, as a means of levels of two variables with the other one variable at its midpoint. The optimum levels of independent variables were elucidated from the response surface plots (Fig. 1d–f) and the desirability profile (Online resource 1c). The experimentally optimized level of independent variables to produce maximum AMP (Y1) by B. velezensis FTL7 was 34.5 °C temperature (X1), 23 h incubation time (X2) and 7.6 pH of the medium (X3). These combinations resulted in an about threefold increase in AMP production by B. velezensis FTL7 as compared to that of the un-optimized conditions. Validation experiments confirmed the aptness of the model equation at the optimized conditions of X1 (34.5 °C), X2 (23.0 h) and 7.6 pH (X3) using different volumes of culture medium (50 mL, 100 mL and 500 mL). The average yield of AMP (13.4 ± 0.2 mm, zone of inhibition against L. monocytogenes Scott A, n = 4) obtained at the optimized conditions significantly (p ≤ 0.05) apt to the value predicted (13.8 mm) by the model equation. These results established the competence and accuracy of the design model for optimizing variables for AMP production by B. velezensis FTL7.

The result of the present study indicates that the AMP production by B. velezensis FTL7 was strongly influenced by different culture conditions, such as incubation temperature, cultivation time and pH of the culture medium (Fig. 1). Incubation temperature is characteristic of a microorganism and greatly affects the synthesis of metabolites and the length of the synthesis stage (Kumar et al. 2021). The incubation period of bioprocess strongly affects the degree of development and production of metabolites (Thadathil et al. 2014; Kumar et al. 2021). The RSM showed the maximum AMP production by B. velezensis FTL7 taking place at a temperature of 34.5 °C. This work showed similarity with the work of Lahiri et al (2021) who showed the maximum production of AMP within a temperature of 30–40 °C. In this study, optimizing culture conditions by a statistical experimental design resulted in an about threefold increase in AMP production by B. velezensis FTL7 (Table 1, Fig. 1b) as compared to that of un-optimized conditions. Suganthi and Mohansrinivasan (2015), in a study, reported a 2.5-fold increase in AMP production by Pediococcus pentosaceus after optimizing certain culture conditions.

Purification of AMPs from B. velezensis FTL7

The whole-genome sequence of the marine bacterium B. velezensis FTL7 was deposited in GenBank with accession no. CP094618 previously (Johny and Suresh 2022). Its complete genome sequence by anti-SMASH and BAGEL-4 revealed the presence of uberolysin/carnocyclin family circular bacteriocin-encoding gene and LP synthetase genes (Johny and Suresh 2022). Further, to identify specific anti-listerial peptides present in the AMPs produced by B. velezensis FTL7, the cell-free supernatant (CFS) was subjected to purification by ammonium sulphate precipitation, butanol extraction and reversed-phase solid phase extraction.

AMPs from B. velezensis FTL7 were produced and partially purified by ammonium sulphate precipitation. The maximum anti-bacterial activity of AMP against L. monocytogenes Scott A (OD₆₀₀ 1.0) was observed with 80% of ammonium sulphate saturation (Fig. 2a). Further butanol extraction was subjected to purify and enrich the AMPs (Fig. 2a). As shown in Fig. 2b, the resultant AMPs fraction exhibited a single band with ~ 6.5 kDa in Tricine-SDS-PAGE analysis. The antibacterial activity of the AMP against L. monocytogenes Scott A was further confirmed by agar-overlay assay (Fig. 2b).

Fig. 2.

Antimicrobial activity of cell-free supernatant (1), ammonium sulphate precipitate (2) and butanol extract (3) of Bacillus velezensis FTL7 against Listeria monocytogenes Scott A (a); tricine-sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE) of purified antimicrobial peptides (AMP). Lane 1: Purified AMP; Lane 2: Standard protein molecular weight marker; Lane 3: Direct overlay of a tricine-SDS-PAGE gel demonstrating the zone of inhibition against L. monocytogenes Scott A (b). ZB = Zobell marine broth

Identification of AMPs from B. velezensis FTL7

MS and MS/MS analysis has demonstrated the advantage in the determination of molecular mass and modifications of a variety of AMPs (Epparti et al. 2022). The C₁₈ solid-phase extraction (SPE) elution with 80% acetonitrile exhibited potent inhibition of L. monocytogenes. MS analysis of this fraction showed several singly and doubly charged ions (Fig. 3a and Table 2). The intact mass was calculated after the deconvolution of ions m/z values and found to be in the range of 1–1.6 kDa. Further, MS/MS of these ions revealed the peptides belonging to three families of LPs-surfactins, -fengycins and -iturin A (Figs. 3, 4a, Table 2 and Online resource 1d-e). The ions observed at m/z 994.63⁺¹, 1008.65⁺¹, 1022.67⁺¹ and 1036.69⁺¹ correspond to the mass of surfactin with fatty acid chain length C₁₂-C₁₅, respectively. Previous studies reported similar ions for surfactin from Bacillus strain (Rungsirivanich et al. 2021). Surfactin is an amphiphilic compound, wherein the heptapeptide sequence forms a lactone ring structure with a chain of β-hydroxy fatty acids (Theatre et al. 2021). MS/MS of ions at m/z 994.63 exhibited fragment ions at m/z 685.44 that arise from the cleavage of the peptide bond by the loss of Glu and C₁₂ β-OH fatty acid. Other fragment ions at m/z 554.36, 441.27, 342.20 and 227.17 indicate the subsequent loss of Leu, Leu, Val and Asp, respectively (Fig. 3b). Thus, the ions at m/z 994.63 confirmed the surfactin sequence to be Glu-Leu-Leu-Val-Asp-Leu-Leu. MS/MS of ions at m/z 1036.69 also comprised fragment ions at m/z 685.44, confirming the loss of Glu from surfactin and C₁₅ β-OH fatty acid (Online resource 1d). Surfactin is known to exhibit emulsification and antimicrobial activities against various pathogens (Rungsirivanich et al. 2021).

Fig. 3.

Mass spectrometry analysis of active antimicrobial peptides (AMPs) fraction indicating the presence of lipopeptides and amylocyclicin (a); MS/MS of singly charged precursor ions at m/z 994.63 matching to surfactin with C₁₂ fatty acid chain (b); MS/MS of doubly charged precursor ions at m/z 529.29 matching to iturin A with C₁₅ fatty acid chain (c); MS/MS of doubly charged precursor ions at m/z 718.38 matching to fengycin A with C₁₄ fatty acid chain (d)

Table 2.

List of lipopeptides produced from Bacillus velezensis FTL7 based on mass spectrometry analysis

S no.	m/z	Charge state	Fatty acid chain length	Sequence
Surfactin
1	994.63/1016.62	[M + H]+/[M + Na]+	C12
2	1008.65/1030.64	[M + H]+/[M + Na]+	C13
3	1022.67/1044.64	[M + H]+/[M + Na]+	C14
4	1036.69/1058.66/1074.64	[M + H]+/[M + Na]+/ [M + K]+	C15
Iturin A
1	1043.55	[M + H]+	C14
2	1057.56	[M + H]+	C15
3	1071.58	[M + H]+	C16
Fengycin A
1	718.38	[M + 2H]2+	C14
2	724.40	[M + 2H]2+	C15
3	725.39	[M + 2H]2+	C15 unsaturated
4	732.40	[M + 2H]2+	C16
5	739.41	[M + 2H]2+	C17
Fengycin B
1	732.40	[M + 2H]2+	C14
2	739.40	[M + 2H]2+	C15
3	746.41	[M + 2H]2+	C16
4	753.42	[M + 2H]2+	C17

Fig. 4.

MS/MS of doubly charged precursor ions at m/z 753.42 matching to fengycin B with C₁₇ fatty acid chain (a); Time-killing curve of purified antimicrobial peptides (AMPs) against Listeria monocytogenes Scott A (b); Scanning electron microscopy (SEM) images of L. monocytogenes cells before (1) and after (2) treatment with AMP (c)

In addition, the precursor ions at m/z 1043.5⁺¹, 1057.50⁺¹/529.26⁺² and 1071.58⁺¹ were attributed to the mass of LP-iturin A having C₁₄-C₁₆ chain fatty acids, respectively. MS/MS of ions at m/z 529.26⁺² indicated that the fragment ions matched to the sequence C₁₅ β-OH fatty acid-Asn-Tyr-Asn-Gln-Pro-Asn-Ser in cyclic form (Fig. 3c). Iturin is a cyclic LP with seven α-amino acid residues and one β-amino acid residue and is produced mainly by species in the B. subtilis group (Shi et al. 2018). Among them, strains of B. velezensis have received much attention for commercial applications and have been studied widely as potential iturinic LP producers (Dunlap et al. 2019).

Fengycins are cyclic lipodecapeptides assembled by lactone ring formation between the carboxylic group of the C-terminal Ile and the hydroxyl group of Tyr (Ongena and Jacques 2008; Daas et al. 2018). Based on the presence of either Ala₆ or Val₆, fengycins are known as fengycin A or fengycin B, respectively. Fengycins contain a saturated or unsaturated β-OH fatty acid with a chain length of C₁₄–C₁₇ at the N-terminus side (Daas et al. 2018). In our analysis, doubly charged precursor ions [M + 2H]²⁺ at m/z 718.38, 725.39, 732.40 and 739.41 matched to the mass of fengycin A variants with fatty acid chains C₁₄, C₁₅, C₁₆ and C₁₇, respectively. Ions at m/z 724.40⁺² were a homolog of C₁₅ fengycin A but with unsaturated fatty acid. In addition, doubly charged ions at m/z 746.41 and 753.42 correspond to fengycin B with C₁₆ and C₁₇ fatty acids, respectively. Further confirmation of fengycin A and B was obtained by MS/MS analysis. Fragmentation of precursor ions at m/z 718.38 indicated an occurrence of ions at m/z 1080.5 and 966.4 arising from the cleavage of Glu-Orn and Orn-Tyr linkages, respectively. The octapeptide ring ion at m/z 966.4 comprised Ala₆ (Fig. 3d). The overall fragmentation pattern confirmed the fengycin A variant with C₁₄-fatty acid. A similar pattern was observed for ions at m/z 732.39 having C₁₆-fatty acid (Online resource 1e). MS/MS of precursor ions at m/z 753.42 revealed fragment ions at m/z 1108.56 and 994.48, representing the cleavage of Glu-Orn and Orn-Tyr linkages, respectively. However, the octapeptide ring ions at m/z 994.48 showed a 28 Da increment compared to the above variants indicating Ala₆ to Val₆ substitution, thus confirming fengycin B (Fig. 4a). Fengycins are attributed to have antibacterial and antifungal activities by creating pores on the cell membrane and membrane distortion on the target cell (Rungsirivanich et al. 2021).

Interestingly, a penta charged ions at m/z 1276.52 [M + 5H]⁵⁺ was also observed in MS analysis (Fig. 3a). Upon deconvolution, the molecular mass matched to amylocyclicin (6377.61 Da) as reported earlier (Scholz et al. 2014; Kurata et al. 2019; Rungsirivanich et al. 2021), which corroborated with Tricine-SDS-PAGE analysis (Figs. 2b and 3a). Due to the higher mass range of these ions, the MS/MS experiment could not be performed. Amylocyclicin exhibits high antibacterial activity against Gram-positive bacteria (Scholz et al. 2014; Kurata et al. 2019; Rungsirivanich et al. 2021). Overall, MS analysis demonstrated that B. velezensis FTL7 could produce two distinct types of AMPs–three variants of LPs and a bacteriocin.

Minimum inhibitory concentration and time-kill kinetics of AMPs

MIC is the lowest concentration at which the AMP completely inhibits the growth of the target microorganism (Kujik et al. 2011). The required MIC of AMP to inhibit the growth of L. monocytogenes was observed as a 2.5 µg/mL concentration of AMP. Furthermore, no visible live cells were observed in the resazurin dye test, which indicated that the AMP had potential bactericidal action (data not shown). Compared with earlier reports on individual peptides from different Bacillus sp., the AMPs from B. velezensis FTL7 displayed high anti-listerial activity (Sebate and Audisio 2013; Liu et al. 2020). It might be due to the combined effect of four types of AMPs from B. velezensis FTL7. The production of more kinds of antibacterial peptides enhances the antibacterial activity of Bacillus sp. and supports their growth in their natural habitat (Falardeau et al. 2013; Liu et al. 2020). MICs for surfactin, fengycins and iturin individually against L. monocytogenes are in the range of 0.125–1 mg/mL, 25 µg/mL and 200–220 µg/mL, respectively (Sebate and Audisio 2013; Lin et al. 2020; Singh et al. 2021). Time-kill kinetics demonstrated the decrease in bacterial load after 30 min treatment using AMP, and the CFU reduced to less than one after 2 h treatment with AMP (Fig. 4b). A comparable observation was reported that treatment of L. monocytogenes cells with JS4/sublichenin reduced the cell count to less than one CFU within 3 h (Wei et al. 2021).

Morphological changes in the membrane of L. monocytogenes Scott A by AMPs treatment

To understand the morphological changes in L. monocytogenes, cells after the treatment with AMP of FTL7 were examined by SEM (Fig. 4c). As shown in Fig. 4c, the control L. monocytogenes cells exhibited a normal cell morphology and shape. However, the AMP-treated L. monocytogenes cells exhibited an abnormal cell shape and morphology with a wrinkled surface, along with adhesion and aggregation of cells (Fig. 4c). It might be due to the leakage of intracellular constituents of the cells by AMP. Similar results have been reported with mejucin (Lee and Chang 2018), bifidocin A (Liu et al. 2017) and subtilin JS-4 (Wei et al. 2021). In a study, Liu et al. (2021) reported the ability of two peptides to reduce the surface hydrophobicity of targeted pathogenic bacterial cells and their mechanisms underlying antimicrobial properties.

Molecular docking studies of AMPs

Automated docking was used to assess the orientation of different LPs (surfactin, fengycin and iturin A) interacting at the binding pocket of internalin, GntR and listeriolysin, respectively (Fig. 5b). The most negative binding energy at zero RMS deviation for each ligand–target interaction was considered the best binding confirmation. The result of the docking studies revealed that fengycin had better interactions with the targets internalin and listeriolysin, while surfactin and iturin A molecules showed better interactions with the GntR (Fig. 5b and Table 3). A high-resolution homology pattern is helpful for solid structural examinations of surface interactions. The 3D folded structure of amylocyclicin was modelled for interaction studies (Fig. 6a). We created the top ten probable binding models using the ZDOCK docking tool and scored based on energy-related scoring. Out of the ten highest-scored internalin–amylocyclicin complexes, the top-scored complex displayed an interface contact area of 1767.797 Å. Similarly, the GntR–amylocyclicin complex showed a contact area of 1086.14 Å, and the listeriolysin–amylocyclicin complex was 876.528 Å (Online resource 1 g and Fig. 6b).

Table 3.

Binding interactions detailing the residues forming hydrogen bond with ligand molecules

Target protein	Ligand molecule	Binding energy (kcal/mol)	Interacting residues
Internalin	Surfactin	− 7.0	THR101, GLN102
	Fengycin	− 7.2	LEU410, ASN413, THR484, THR485, ASP414
	Iturin A	− 6.2	ASN413, THR484, THR485, LEU410, GLN409
GntR	Surfactin	− 6.4	TYR10, LYS11, ARG51, ASP85,
	Fengycin	− 6.3	HIS15, HIS18, GLN89, ASN93, ASP4, PRO8
	Iturin A	− 5.7	LYS11, HIS15, HIS18, ARG78, LS88, GLN89, ASP85, GLN89, GLU55
Listeriolysin	Surfactin	− 6.0	LYS482, TYR520, ASP498, SER521
	Fengycin	− 6.1	ASP497, THR494, APS498, TYR520, ARG493
	Iturin A	− 5.8	ASN457, TYR480, LYS482, THR494

Fig. 6.

Three-dimensional folded structure of amylocyclicin (a); interaction of amylocyclicin with internalin (A), GntR (B), listeriolysin (C) (b)

Theoretically, computational analysis is considered one of the best approaches for exploring the mechanism of interaction between a protein and its ligand. In the present study, the effective interactions of B. velezensis FTL7 AMPs with internalin might be attributed to their anti-listerial effect correlated with the L. monocytogenes adhesion and invasion to human epithelial cells (Sánchez-González et al. 2022). In line with our findings, earlier studies have revealed the mechanisms involving the inhibition of internalin associated with the potential glutamic acid displacement (Sánchez-González et al. 2022), inactivation of the srtA gene (Bierne et al. 2002), inhibition of E-cadherin, the receptor for internalin (Sánchez-González et al. 2022), etc. Further, the inhibition of the virulence protein, listeriolysin, representing an alternative strategy for fighting L. monocytogenes infection, was found to be effective by natural AMPs. Listeriolysin is a cholesterol-dependant cytolysin that helps L. monocytogenes to escape from phagosomes and allows the bacteria to multiply inside the host (Cheng et al. 2020). Our results indicate that AMPs from FTL7 have better interaction with listeriolysin, and internalin may represent a promising anti-listerial agent. GntR can also be a potential target as this transcription factor family is highly conserved in the bacterial kingdom. GntR-family transcriptional regulators are responsible for developing L. monocytogenes biofilms, which contributes to persistent contamination during food processing. L. monocytogenes Scott A is the strong biofilm-forming strain, and the GntR-family regulators were highly expressed in Scott A (Wassinger et al. 2013). Our studies also suggest that AMPs from B. velezensis FTL7 strongly interact with GntR-family transcriptional regulators.

Conclusion

In this investigation, the optimization of cultural conditions by a statistical experimental design resulted in an about threefold increase in AMP production by B. velezensis FTL7 as compared to that of the un-optimized conditions. The purified AMPs from B. velezensis FTL7 were identified as surfactins, fengycins, iturin A and amylocyclicin by the MS and MS/MS analysis. The minimum inhibitory concentration of these AMPs against L. monocytogenes Scott A was 2.5 µg/mL. The AMPs from B. velezensis FTL7 inhibited the growth of L. monocytogenes Scott A by distressing the cell integrity and membrane permeability. In-silico docking studies showed that the AMPs from B. velezensis FTL7 have high binding energy and stable binding patterns towards L. monocytogenes target proteins and confirmed the effectiveness of these AMPs as anti-listerial. All these findings suggest that the AMPs from B. velezensis FTL7 have great potential as a natural food preservative to control L. monocytogenes.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

LCJ: investigation, data curation and writing––original draft. BSGK: data analysis and writing––review and editing. SJAR: in-silico studies and review and editing. PVS: conceptualization, supervision and writing––review and editing.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

Data availability

Available from the corresponding author on reasonable request.

Declarations

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.