Bioprotective Effect of Lactococcus piscium CNCM I-4031 Against Listeria monocytogenes Growth and Virulence

Listeria monocytogenes is a Gram-positive pathogen occurring in many refrigerated ready-to-eat foods. It is responsible for foodborne listeriosis, a rare but severe disease with a high mortality rate (20–30%). Lactococcus piscium CNCM I-4031 has the capacity to prevent the growth of L. monocytogenes in contaminated peeled and cooked shrimp and in a chemically defined medium using a cell-to-cell contact-dependent mechanism. To characterize this inhibition further, the effect of L. piscium was tested on a collection of 42 L. monocytogenes strains. All strains were inhibited but had different sensitivities. The effect of the initial concentration of the protective and the target bacteria revealed that the inhibition always occurred when L. piscium had reached its maximum population density, whatever the initial concentration of the protective bacteria. Viewed by scanning electron microscopy, L. monocytogenes cell shape and surface appeared modified in co-culture with L. piscium CNCM I-4031. Lastly, L. monocytogenes virulence, evaluated by a plaque-forming assay on the HT-29 cell line, was reduced after cell pre-treatment by the protective bacteria. In conclusion, the bioprotective effect of L. piscium toward L. monocytogenes growth and virulence was demonstrated, and a hypothesis for the inhibition mechanism is put forward.


INTRODUCTION
Listeria monocytogenes is a human pathogenic Gram-positive bacterium, which is responsible for foodborne listeriosis generally associated with a high mortality rate (20-30%). All human population groups can be infected and particularly newborn infants, pregnant women, elderly people, and immuno-compromised patients (Lecuit et al., 2015). This species constitutes a major problem in refrigerated ready-to-eat (RTE) foods (Rocourt et al., 2003). According to the Codex Alimentarius Commission, RTE products are "any food which is normally eaten in its raw state or any food handled, processed, mixed, cooked, or otherwise prepared into a form which is normally eaten without further listericidal steps" (CAC, 2007). L. monocytogenes differs from most other food-borne pathogens in that it is ubiquitous and can grow or survive in most conditions encountered in the food chain and food-processing procedures (Buchanan et al., 2017). Among foodstuffs, fishery products recorded the highest level of non-compliance with EU safety criteria in 2015 (EFSA, 2016) while seafood products were implicated in 15% of L. monocytogenes outbreaks reported in Europe in the last 3 years. Its survival ability makes the control of this microorganism in food products, especially RTE foods, a major challenge. The lactic acid bacteria (LAB), generally recognized as safe (Salminen et al., 1998), are also present and dominant in RTE foods such as meat and seafood products stored under vacuum or modified atmosphere packaging. Selected LAB can limit the development of L. monocytogenes and are thus recognized as efficient bioprotective agents in food systems (Brillet et al., 2004;Nilsson et al., 2005;Tomé et al., 2006;Vermeiren et al., 2006a;Unlu et al., 2015). The competition between LAB and L. monocytogenes involves various bactericidal or bacteriostatic mechanisms such as (i) competition for nutrients (Nilsson et al., 2005); (ii) production of one or more antimicrobial active metabolites such as bacteriocins (Richard et al., 2003;Dortu et al., 2008;Martinez et al., 2015), reuterin (El-Ziney et al., 1999), organic acids (Amezquita and Brashears, 2002), and hydrogen peroxide (Ito et al., 2003;Batdorj et al., 2007). Lactococcus piscium CNCM I-4031 is an efficient bioprotective strain for seafood products isolated from raw salmon stored under modified atmosphere packaging (Matamoros et al., 2009a). It improves the sensory quality of cooked shrimp by preventing the growth of Brochothrix thermosphacta (Fall et al., 2012). L. piscium CNCM I-4031 can also limit the growth of L. monocytogenes RF191 during the storage of cooked shrimp (Fall et al., 2010). The inhibition mechanism has not yet been entirely elucidated but using a chemically defined medium (MSMA) to mimic the shrimp matrix, it has been suggested that cell contact is required for the inhibition of L. monocytogenes RF191 by L. piscium CNCM I-4031 (Saraoui et al., 2016).
The aim of this study was to characterize the further inhibitory effect of L. piscium CNCM I-4031 against L. monocytogenes species. First, the inhibitory activity of the CNCM I-4031 strain was evaluated on a collection of 42 L. monocytogenes strains. Then, the effect of the protective strain CNCM I-4031 on the growth, morphological shape, and virulence of L. monocytogenes was investigated.

Bacterial Strains, Culture Media, and Conditions
Lactococcus piscium CNCM-I 4031 was isolated from fresh salmon steak packed under modified atmosphere packaging (Matamoros et al., 2009b). The L. monocytogenes strains used in this study are listed in Table 1. All strains were stored in aliquots of 250 µl at −80 • C in a final concentration of 10% (v/v) of glycerol. For all experiments, an aliquot of the strain was subcultured in Elliker broth (Biokar Diagnostic, Beauvais, France) for 24 h at 26 • C for L. piscium, and in Brain Heart Infusion supplemented by 2% NaCl (mBHI) (Biokar Diagnostic, Beauvais, France) for 24 h at 20 • C for L. monocytogenes. The cultures were diluted in their culture medium to obtain appropriate initial cell concentrations. The chemically defined medium MSMA used for the bacterial interaction observation  was prepared as previously described by Saraoui et al. (2016). L. piscium was enumerated by spreading 100 µl of 10-fold serial dilutions on Elliker agar plates incubated at 8 • C for 5 days under anaerobiosis (co-cultures) or at 26 • C for 48 h (pure cultures). L. monocytogenes was enumerated by spread-plating 100 µl (classic enumeration) or 5 µl (microenumeration method) of 10-fold serial dilutions on mBHI agar incubated at 37 • C for 24 h.
Biodiversity of L. monocytogenes Sensitivity to L. piscium CNCM I-4031 After subculture, L. piscium CNCM I-4031 and each of the 42 L. monocytogenes strains ( Table 1) were co-cultured in a 96-well microplate filled with 200 µl of MSMA at an initial concentration of 10 6 and 10 3 CFU/ml, respectively. The microplate was incubated at 26 • C for 30 h without shaking. Controls consisted of monocultures of each L. monocytogenes strain inoculated in the same conditions. L. monocytogenes strains were enumerated in CFU/ml using the microenumeration method described in section "Bacterial Strains, Culture Media, and Conditions." The inhibition was calculated by the difference in the log-concentration of L. monocytogenes strain in pure culture and in co-culture. The standard deviation was estimated according to five independent replicates of the inhibition tests for L. monocytogenes RF191 and L. piscium CNCM-4031.
Effect of Co-culture Ratios on the Inhibition of L. monocytogenes by L. piscium CNCM I-4031 Lactococcus piscium CNCM I-4031 and L. monocytogenes RF191 were co-inoculated in 250-ml flasks of MSMA medium, without shaking. Depending on the experiments, the initial ratios between L. piscium and L. monocytogenes were 10 3 /10 3 (A), 10 5 /10 3 (B), and 10 6 /10 6 (C) CFU/ml, respectively. The co-cultures were incubated at 26 • C, and the growth of both strains was monitored during 48 to 72 h by the classic enumeration method described in section "Bacterial Strains, Culture Media, and Conditions." Controls consisted of monocultures of each strain in MSMA at 26 • C at the same inoculation levels. All the cultures were performed in triplicate.

Observation of Cells in Co-culture by Scanning Electron Microscopy
Pure cultures and co-cultures (ratio 10 6 /10 3 CFU/ml) of L. piscium CNCM I-4031 and L. monocytogenes RF191 were cultivated in 10 ml of MSMA medium at 26 • C for 24 h. Then, 1 ml (10 8 cells) of the suspension was filtered on a Nuclepore R polycarbonate membrane with a 0.22-µm pore size and 13-mm diameter (Whatman International Ltd., Maidstone, United Kingdom). In order to observe the bacterial interaction on a solid medium (Dubey and Ben-Yehuda, 2011), another filter membrane was placed on an MSMA agar (15 g/l) plate and spotted with 10 µl of co-culture and incubated for 6 h at 26 • C. were routinely grown in 75-cm 2 flasks (Sigma) in a complete medium, DMEM (Dulbecco's modified Eagle's medium), with 10% (v/v) fetal calf serum (SCF), and 1% amphotericin B 250 µg/ml (Eurobio). Gentamicin 100 µg/ml (Sigma) was added to the culture medium. Cells were kept in the humidified atmosphere of a 5% CO 2 incubator at 37 • C. One hundred microliter of HT-29 cell suspension (2 × 10 4 cells) was deposited per well in a 96-well tissue culture plate (Sigma). The plates were incubated for 4 days with antibiotics followed by incubation for 24 h without antibiotics to obtain confluent monolayers.
In each well, HT-29 cells were infected with 10 8 L. piscium CNCM I-4031 and incubated for 1 h at 37 • C. L. piscium cells were then removed by suction, and the HT-29 cells were infected with 100-µl suspensions from 10 8 to 10 3 CFU/ml of L. monocytogenes per well and incubated for 2 h at 37 • C. L. monocytogenes cells were then removed by suction. The HT-29 cells were covered with 100 µl of DMEM 10% SCF with 100 µg/ml of gentamicin and incubated for 1.5 h at 37 • C to eliminate bacterial cells from the plates. Each well was then overlaid with DMEM 10% SCF with 100 µg/ml of gentamicin containing 0.47% of indubiose to prevent cell starvation. Incubation was carried out at 37 • C for 24 to 48 h. The virulence of the bacterial cells was evaluated by plaque-forming assay (PFA) in the cell monolayers using an optical microscope (VWR, Pennsylvania, United States). Controls consisted of HT-29 cells infected with serial dilutions of 10 8 to 10 2 UFC of L. piscium CNCM I-4031 or L. monocytogenes RF191 or ScottA per well. The whole experiment was repeated three times (different weeks) corresponding to nine independent tests.

Statistical Analyses
Statistical analyses on bacterial counts concentrations were performed using R software (R Core Team, 2014) by the analysis of one-way analysis of factor variance (ANOVA) followed by least significant difference (LSD) test.

RESULTS AND DISCUSSION
Inhibition Capacity of L. piscium CNCM I-4031 Strain Toward Various L. monocytogenes Strains A total of 42 L. monocytogenes strains from diverse sources, and geographical origins were selected ( Table 1). Most of them were isolated from seafood except the reference strains ScottA (Fleming et al., 1985) and EGD-e (Murray et al., 1926) isolated during a listeriosis outbreak from human and animal tissue, respectively. The inhibition of L. monocytogenes strains was measured after 30 h of co-culture on microplates with L. piscium CNCM I-4031, in MSMA liquid medium. In these miniaturized experimental conditions, L. monocytogenes RF191 displayed an inhibition of 2.40 ± 0.35 log CFU/ml, which is lower than that previously described in larger culture volumes with an inhibition of 3-4 log CFU/ml (Saraoui et al., 2016). However, this screening test showed that L. piscium CNCM I-4031 could inhibit all L. monocytogenes strains, whatever their origin (Figure 1). The inhibition yield was straindependent, varying between 0.69 log CFU/ml for EU2162 and 3.72 log CFU/ml for EU2158. These values can be correlated with the low growth in MSMA medium of EU2262 in pure culture (7.3 log CFU/ml) and, in contrast, the strong growth of EU2158 (9.3 log CFU/ml). In mixed culture, after 30 h of incubation, the concentration of L. monocytogenes ranged between 5.6 and 7.1 log CFU/ml when cultivated alone and 7.3 and 9.5 in coculture (Figure 2).

Influence of the Initial Ratio of L. piscium CNCM I-4031/L. monocytogenes
Previous studies have shown that L. piscium CNCM I-4031 can inhibit the growth of L. monocytogenes RF191 from 3 to 4 log units with an inoculum ratio of 10 6 /10 3 CFU/ml (L. piscium/L. monocytogenes) in shrimp at 8 • C and in MSMA medium at 26 • C (Fall et al., 2010;Saraoui et al., 2016). Considering that the concentration of pathogenic bacteria in food at the beginning of storage is low, protective bacteria are usually added at high concentrations to food products (Ananou et al., 2005;Brillet et al., 2005). In order to determine whether the inhibition was linked with these initial concentrations, three different initial ratios of L. piscium/L. monocytogenes RF191 were analyzed (Figure 3): 10 3 /10 3 CFU/ml (A), 10 5 /10 3 CFU/ml (B), and 10 6 /10 6 CFU/ml (C).
The growth kinetics of L. piscium in pure and co-culture were similar whatever the initial ratios. The maximum population density (MPD) of approximately 10 8 CFU/ml was reached after 40-48 h in (A), 25 h in (B), and 15 h in (C) with a growth rate between 0.32 to 0.45 h −1 ( Table 2). L. monocytogenes growth started with an exponential phase and reached the MPD (∼10 9 CFU/ml) at 30 h in (A), 20 h in (B), and 15 h in (C) ( Table 2). In each co-culture, the L. monocytogenes population increased to reach 10 8 to 10 9 CFU/ml until L. piscium achieved the MPD, then decreased (Figure 3). These results indicate that, regardless of the inoculum ratio of the two strains, L. monocytogenes was always significantly inhibited in co-cultures when L. piscium reached its MPD. This inhibition was proportional to the initial concentration of L. piscium and ranged from 1.42 to 3.37 log CFU/g ( Table 2) with a higher inhibitory effect when the protective strain was inoculated at 10 5 -10 6 CFU/ml. These observations are in accordance with previous studies with other bioprotective LAB isolated from seafood, such as a non-bacteriocin-producing Carnobacterium piscicola A9b (Nilsson et al., 2005) and Lactobacillus sakei 10A (Vermeiren et al., 2006b). Nevertheless, contrary to L. sakei 10A, which required an initial concentration of up to 10 5 CFU/g to inhibit L. monocytogenes effectively, L. piscium showed an inhibition with a low initial concentration of 10 3 CFU/ml.
The competition between LAB species and other populations in food and mixed cultures by one single "dominant" strain and when LAB have reached their maximum level is described in the literature as the "Jameson effect" (Jameson, 1962;Gimenez and Dalgaard, 2004). The Jameson effect is considered a race between species in order to maximize their growth by exploiting the environmental nutrients. Thus, the species that first reaches its MPD inhibits the growth of the other species (Mellefont et al., 2008;Cornu et al., 2011). However, in our experiments, we showed that the inhibition of L. monocytogenes by L. piscium occurred even when the concentration of the pathogenic bacteria was higher than that of the protective bacteria (Figures 3A,C). In these conditions, L. monocytogenes reached its MPD before L. piscium but no inhibition of L. piscium occurred. The same results were found when L. monocytogenes ScottA was co-inoculated with Escherichia coli whereas in co-culture with Lactobacillus plantarum or Pseudomonas fluorescens, the first strain reaching its MPD stopped the growth of the other one (Mellefont et al., 2008).
In a previous study, we demonstrated that the amount of lactic acid produced by L. piscium CNCM I-4031 in co-culture conditions was not responsible for the inhibition (Saraoui et al., 2016). Moreover, the supplementation of the co-culture with nutrients did not restore the ability of L. monocytogenes to grow, and no inhibition was observed when both cultures were separated by a 0.45-µm membrane. The present study evidenced that L. piscium could limit the growth of L. monocytogenes when it reached its maximum density, whatever its initial level.
Inhibition phenomena linked to maximum cellular concentration have been shown in other LAB, and some authors have suggested that they could involve quorum sensing (Kuipers et al., 2000;Risøen et al., 2000;Rohde and Quadri, 2006;Moslehi-Jenabian et al., 2011;Rizzello et al., 2014). In our case, the mechanism remains unknown and has still to be investigated.

Observation of Cells in Co-culture by Scanning Electron Microscopy
With the aim of investigating the behavior of the pathogenic bacteria in co-culture at the microscopic scale, the cell morphology of L. piscium CNCM I-4031 and L. monocytogenes RF191 was compared in monoculture in MSMA medium with co-culture conditions at the time of inhibition. The results presented in Figure 4 show that L. piscium in monocultures were spherical (Elliker) or ovoid (MSMA) cells between 0.5 and 1 µm in diameter and appeared in pairs or short chains. L. monocytogenes were rod-shaped cells between 0.5 µm (MSMA) and 1.5 µm (mBHI) in length and ∼0.5 µm in width, appearing individually or in pairs ( Figure 4B). For co-cultures, strains were observed in either liquid or solid media (Figures 4C,D). Co-cultures on solid media have already revealed the presence of nanotubes, which are extensions of the cytoplasmic membrane used by bacteria to exchange their cellular compounds in cell contact interactions (Dubey and Ben-Yehuda, 2011). These authors described intra-species nanotubes between cells of Bacillus subtilis, as well as interspecies connections between cells of B. subtilis, S. aureus, and E. coli using scanning electron microscopy. In the conditions of our experiments, such nanotubes were not observed between the cells. However, the pictures show that in the presence of L. piscium, L. monocytogenes cells appeared to be more elongated (Figures 4C,D). The elongation of Listeria cells in stress conditions (low pH, high salt concentration) has been reported in previous studies (Besnard et al., 2000;Bereksi et al., 2002). Listeria cells can divide without septation, leading to the modification of their surface properties with the presence of filamentous structures. Our observations also suggest that the surface of L. monocytogenes cells was damaged or completely altered in the presence of LAB (Figures 4C,D). These effects on L. monocytogenes morphology have been reported in previous studies dealing with the mechanism of action of anti-Listeria components (Dieuleveux et al., 1998).
Effect of L. piscium CNCM I-4031 on the Virulence of L. monocytogenes As an important pathogen, L. monocytogenes encompasses a large spectrum of strains with varying virulence effects (Buchanan et al., 2017). In order to examine another aspect of the bioprotective effect of L. piscium, the impact on L. monocytogenes virulence was investigated. For this purpose, the virulence of two selected L. monocytogenes strains, RF191 isolated from seafood and ScottA known to be a highly virulent strain (Lindback et al., 2010), were tested using an HT-29 cell PFA, in the presence or absence of L. piscium.
On an HT-29 monolayer pre-treated with 10 8 L. piscium, no lysis plates were detected after 24 or 48 h confirming that L. piscium has no pathogenic activity. Plaques were counted in the well containing HT-29 cells infected by 5 log CFU/ml of L. monocytogenes ScottA or RF191 strains. The ScottA strain formed large deep plaques whereas the RF191 strain formed small shallow ones. The mean log PFA values were 3.93 ± 0.06 and 3.68 ± 0.08, respectively (Figure 5). According to the study of Roche et al. (2001), the RF191 strain that forms more than 3.34 log plaques should be considered virulent, even though it remains less virulent than the ScottA strain. When HT-29 cells were infected by L. monocytogenes strains after being treated by L. piscium, the mean log PFA values were 3.40 ± 0.36 and 1.25 ± 0.29 for ScottA and RF191 strains, respectively (Figure 5). A significant effect was found for the RF191 strain that had lost its virulence after pretreatment with L. piscium CNCM I-64031. A slight but a significant decrease in virulence was also observed for the ScottA strain, suggesting that the effect of the protective bacteria on L. monocytogenes virulence is strain-dependent. The effect on virulence reduction by LAB has already been described for foodborne pathogens such as L. monocytogenes (Garriga et al., 2015;Pilchova et al., 2016) or Campylobacter (Alemka et al., 2010). The effect is usually investigated as one of the probiotic properties of bacteria; however, such additional properties also increase the bioprotective value of strains and their safety assessment.

CONCLUSION
In this study, we have demonstrated the ability of L. piscium CNCM I-4031 to inhibit a large collection of L. monocytogenes strains. This inhibition occurs whatever the initial concentration of the protective strain when L. piscium has reached its MPD. The inhibition mechanism is still under investigation but our different sets of results suggest that it requires the proximity of cells and affects the cellular surface of the targeted bacteria. In addition to inhibiting L. monocytogenes growth, L. piscium decreased L. monocytogenes virulence with a variable effect according to the strain. This study provides additional knowledge about the inhibitory activity of a nonbacteriocin-producing LAB toward L. monocytogenes and significant information for the potential use of L. piscium CNCM I-4031 to control L. monocytogenes in seafood products.

AUTHOR CONTRIBUTIONS
M-FP and FL designed the work. TS, DP, and FC performed the experiments. TS and FL assisted with the scanning electron microscopy analysis. TS, M-FP, FL, DP, and J-MC analyzed the data. TS drafted the paper. MF-P, FL, and DP wrote the final version of the manuscript, which was read and approved by all authors.

FUNDING
TS was the recipient of a Ph.D. fellowship from the French Ministry of Higher Education and Research. This work was financially supported by the project COM-BACT from the Région des Pays de la Loire, France.