Inhibition of Bacterial Growth, Enterotoxin Production, and Spore Outgrowth in Strains of Bacillus cereus by Bacteriocin AS-48

Abstract

Bacteriocin AS-48 showed high bactericidal activity for mesophilic and psychrotrophic strains of Bacillus cereus over a broad pH range. AS-48 inhibition of the enterotoxin-producing strain LWL1 was enhanced by sodium nitrite, sodium lactate, and sodium chloride. The latter also enhanced AS-48 activity against strain CECT 131. Bacterial growth and enterotoxin production by strain LWL1 were completely inhibited at bacteriocin concentrations of 7.5 μg/ml. At subinhibitory bacteriocin concentrations, enterotoxin production decreased markedly and sporulation was delayed. Intact spores were resistant to AS-48 but became gradually sensitive to AS-48 during the course of germination.

Bacillus cereus is an aerobic spore-forming bacterium that is widely distributed in the environment, mainly in soil, from which it is easily spread to many types of foods, especially those of vegetable origin, as well as meat, eggs, milk, and dairy products (2, 26, 31). This bacterium is one of the leading causes of food poisoning in the industrialized world, causing gastrointestinal disorders (19, 26). B. cereus produces one emetic toxin and at least three different enterotoxins (3, 19, 27), which are responsible for separate emetic and diarrheal syndromes, respectively. The emetic response has been associated with consumption of rice, pasta, pastry, and noodles in which B. cereus has grown and produced the toxin (27, 37). Diarrheal outbreaks are caused by the enterotoxins produced during vegetative growth of the bacterium in the small intestine, and a wide variety of foods, including meat and vegetable dishes, soups, and dairy products, have been implicated in such outbreaks (16). Furthermore, observations that toxin-producing psychrotrophic strains have been implicated in outbreaks of food-borne illness (9, 10, 18, 21) have raised concerns about their growth and toxin production in refrigerated foods.

The use of bacteriocins either alone or in combination with physicochemical treatments to arrest spore outgrowth and enterotoxin production may be an efficient way to prevent B. cereus food poisoning (23). Bacteriocin AS-48 is a cationic cyclic peptide produced by Enterococcus faecalis S-48 (11, 12, 34). Because of its broad spectrum of antimicrobial activity (1, 13, 14, 29), its stability at a wide range of temperatures and pH values, and its sensitivity to digestive proteases (11, 12, 34), AS-48 is a promising candidate for food biopreservation. In the present work, we studied the activity of AS-48 against vegetative cells and spores of mesophilic and psychrotrophic enterotoxigenic B. cereus strains, as well as the influence of environmental factors (refrigeration temperature, pH, and chemical preservatives) in foods on the antimicrobial effectiveness of AS-48.

E. faecalis A-48-32 and S-47 (from our collection) were used as bacteriocin producer and indicator strains, respectively. Bacteriocin AS-48 was produced in buffered CM-G (11) and then purified to homogeneity by cation-exchange chromatography on carboxymethyl Sephadex CM-25, followed by reversed-phase high-performance liquid chromatography (12). A single batch of bacteriocin was used for this study. The protein concentration was determined as described previously (7). Bacteriocin activity was titrated by the agar well diffusion assay, using wells that were 8 mm in diameter (13). The activity of AS-48 was tested with mesophilic, nonenterotoxigenic strains and psychrotrophic, enterotoxigenic strains that produce diarrheic toxin (strains LWL1, LWL3, and LWL10) (10) (Table 1). The mesophilic and psychrotrophic strains were grown at 37°C. All of the strains, especially the enterotoxigenic strains, were very sensitive to AS-48 (Table 1). Strains CECT 131 and LWL1 were used for a more detailed study.

TABLE 1.

Antimicrobial activity of bacteriocin AS-48 against some mesophilic and psychrotrophic strains of B. cereus, as well as the indicator strain E. faecalis S-47

Strain	Halo of inhibition diam (mm)a
Mesophilic strains
B. cereus CECT 131	16 ± 0.5
B. cereus CECT 495	13 ± 1.0
B. cereus CECT 4014	19 ± 0.5
Psychrotrophic strains
B. cereus LWL1	19 ± 0.5
B. cereus LWL3	19 ± 0.5
B. cereus LWL10	19 ± 0.5
E. faecalis S-47	21 ± 1.0

^aSamples were tested by using the agar well diffusion assay and 5 μg per well. The diameter of the halo of inhibition includes the well diameter (8 mm). The values are averages ± standard deviations based on three independent experiments.

The effect of AS-48 on vegetative cells at different temperatures was studied by using exponential-phase cultures of strains CECT 131 and LWL1 grown in brain heart infusion broth (BHI) (Oxoid, Basingstoke, United Kingdom) to an optical density at 620 nm (OD₆₂₀) of 0.1 (ca. 10⁷ CFU/ml). Then AS-48 was added at final concentrations of 5 and 10 μg/ml. At intervals during incubation, samples obtained from controls and treated cultures were serially diluted in a sterile saline solution (0.85% NaCl) and plated in triplicate on tryptic soy agar (AdsaMicro, Barcelona, Spain) to determine the number of viable cells. The plates were incubated at 37°C for 48 h, and the number of colonies was determined in order to calculate the average number of CFU for three independent experiments. The antimicrobial activity of AS-48 against strains CECT 131 and LWL1 was influenced by incubation temperature. At 37°C a bacteriocin concentration of 5 μg/ml caused a marked decrease in the initial numbers of viable cells of both strains within the first 1 h of incubation, but then cultures resumed exponential growth (Fig. 1A). In the presence of 10 μg of AS-48 per ml growth was totally inhibited, and no viable cells were detected in 0.1-ml portions of cultures after 4 h of incubation for strain CECT 131 or after 24 h of incubation for strain LWL1. For activity assays at 5 and 15°C, the cells from exponential-phase cultures were precooled for 1 h at the desired temperature and incubated with different concentrations of AS-48 for 1, 24, 48, and 72 h. When 15°C and a bacteriocin concentration of 5 μg/ml were used, cultures of strain LWL1 recovered more slowly (48 to 72 h), while cultures of strain CECT 131 were almost totally inhibited (Fig. 1B). When 10 μg of AS-48 per ml was used, a few viable cells (10 to 100 CFU/ml) were still detected after 72 h of incubation for both strains. At 5°C, the numbers of viable cells of both strains were markedly reduced within the first 24 h during incubation with AS-48, and further growth was totally inhibited by 5 μg of AS-48 per ml (Fig. 1C). Most probably, the greater efficacy of AS-48 under these conditions could be attributed to a slowing of growth caused by the low temperature, which resulted in a higher number of bacteriocin molecules per cell as well as longer exposure of sublethally injured cells to the damaging bacteriocin. The sizes of the surviving populations (which represented less than 0.01% of the initial populations) decreased very slowly during the subsequent 48 h of incubation. These populations probably consisted of nongerminated spores present in the initial inoculum, because intact spores were resistant to AS-48. In fact, when cultures treated with AS-48 for 72 h were heat shocked to induce spore germination, the remaining viable cells were unable to grow on plates containing AS-48 (5 μg/ml). At 5°C the effects of 5 and 10 μg of AS-48 per ml on strain LWL1 were almost identical, suggesting that at a low temperature the effect of AS-48 depends more on the cell conditions than on the bacteriocin concentration.

FIG. 1.

Effect of AS-48 on the numbers of viable cells in exponential-phase cultures of B. cereus CECT 131 (open symbols) and B. cereus LWL1 (solid symbols). Cultures were incubated at 37°C (A), 15°C (B), or 5°C (C) with no bacteriocin (circles), 5 μg of AS-48 per ml (squares), or 10 μg of AS-48 per ml (triangles).

When cells of strain LWL1 adapted to a low temperature by incubation for 4 weeks at 5°C were incubated at 5°C with AS-48 for 6 h, no decrease in viable counts was observed in the presence of final bacteriocin concentrations up to 50 μg/ml. These results agree with those obtained for Listeria monocytogenes (29). Psychrotrophic microorganisms frequently adapt to cold environments by changing their membrane lipid composition to maintain membrane fluidity (4). Such changes are expected to have large effects on sensitivity to antimicrobial substances (like AS-48) that act on the bacterial cytoplasmic membrane. Therefore, we investigated the membrane fatty acid composition of strain LWL1. Cells were grown in Penassay broth at 37 and 5°C to an OD₆₂₀ of 0.4 as described by Kaneda (25) and were washed with peptone water. Lipids were extracted (6, 35, 38), and the fatty acid methyl esters were analyzed as described previously (29) by using iso and anteiso fatty acid standards (Sigma). In cold-adapted B. cereus LWL1 there were increases in the proportions of tetradecanoic acid and the branched fatty acid 11-methyllaurate and a decrease in the proportion of hexadecanoic acid (Table 2). The latter compound was the predominating fatty acid at 37°C. These changes resemble those observed in L. monocytogenes, in which the amount of straight fatty acids decreased upon incubation at a low temperature and the proportion of branched C₁₅ fatty acid increased (28). Strains of L. monocytogenes adapted to AS-48 also had modified membrane lipid compositions (with increased proportions of branched fatty acids, especially C₁₅ and C₁₇, and much lower proportions of saturated C₁₆ fatty acids), which increased membrane fluidity (29). Thus, these modifications seem to play a role in the increased resistance to AS-48 not only in AS-48-adapted listerial strains but also in cold-adapted B. cereus LWL1. Recently, it has been proposed that bacteriocin AS-48 may act by a molecular electroporation mechanism in which the local accumulation of positive charges carried by AS-48 induces disruption of the membrane curvature (17). Therefore, further studies should be carried out to interpret the role of membrane fatty acid composition in the bacteriocin-membrane interaction.

TABLE 2.

Fatty acid composition of B. cereus LWL1 grown at different temperatures

Fatty acid	% at 37°Ca	% at 5°Ca
Dodecanoic acid	0.22 ± 0.41	1.12 ± 0.25
11-Methyllauric acid	0.21 ± 0.34	12.14 ± 0.74
Tridecanoic acid	0.12 ± 0.15	1.42 ± 0.56
Tetradecanoic acid	1.70 ± 0.45	16.74 ± 1.02
13-Methylmyristic acid	3.18 ± 0.87	0.25 ± 0.1
Hexadecanoic acid	70.28 ± 1.38	37.73 ± 0.87
Octadecanoic acid	24.83 ± 0.74	30.78 ± 1.23

^aThe values are averages ± standard deviations based on three independent experiments.

The influence of pH on the antimicrobial activity of AS-48 against B. cereus CECT 131 was also studied. Cells from an overnight culture in BHI were collected by centrifugation, washed twice with the sterile saline solution, and resuspended in buffer (0.1 M orthophosphoric acid-NaOH [pH 5] or 0.1 M sodium phosphate [pH 6 to 8]) to an OD₆₂₀ of 0.1. Bacterial cell suspensions that were supplemented or not supplemented with AS-48 (5 μg/ml) were incubated at 37°C for 1 h and then serially diluted and plated to determine the number of remaining viable cells. The reductions in the number of viable cells were approximately 2.5 log units for pH 5 and approximately 3.5 log units for pH 6, 7, and 8. The increase in activity observed at a neutral to alkaline pH could be useful for protecting food against B. cereus, because this bacterium can grow at acidic pH values and raise the medium pH to neutral (23). The antimicrobial activities of many chemical preservatives (i.e., organic acids) and bacteriocins, such as nisin, sakacin P, and curvacin A, are enhanced at low pH values but are much lower at neutral and alkaline pH values (15, 23). Therefore, it is of great interest to find substances that can be used to preserve foods that cannot be acidified.

The chemical preservatives sodium nitrite (150 ppm), sodium lactate (0.5%), sodium benzoate (0.1%), sodium sorbate (0.2%), and sodium chloride (3 and 5%) were tested at neutral pH with exponentially growing cultures (OD₆₂₀, 0.1) at 37°C either alone or in combination with AS-48 (5 and 10 μg/ml). None of the organic acids tested alone had any antimicrobial effect. Most organic acids have pK_a values of 3 to 5 and therefore require pH values of less than 5.5 to be effective (8). Sodium benzoate and sodium sorbate had no effect on cultures of strains CECT 131 and LWL1 treated with AS-48. Sodium nitrite and sodium lactate had no effect on growth of either strain, but they enhanced the activity of AS-48 against strain LWL1, as shown by the lower bacteriocin concentration (5 μg/ml) required for effective inhibition (Fig. 2A and B). These results are very important because sodium lactate is present in all fermented foods. Nevertheless, AS-48 activity against strain CECT 131 was not potentiated or inhibited by the organic acids tested (data not shown). The antibacterial activity of nitrite is also enhanced greatly by an acidic pH, although strains of Bacillus are resistant to this compound (20). A synergistic inhibitory effect of nitrite and nisin on outgrowth of Clostridium sporogenes spores has also been reported to depend on an acidic pH (33). The activity of AS-48 against both strains of B. cereus tested was enhanced by sodium chloride, like the activity of other bacteriocins, such as curvacin A and especially sakacin P, against Lactobacillus curvatus and Listeria innocua (15). The numbers of viable cells after 24 h of incubation with AS-48 (5 μg/ml) plus 5% NaCl were more than 5 log units lower for strain LWL1 (Fig. 2C) and more than 3 log units lower for strain CECT 131 (Fig. 2D) than the numbers of viable cells in cultures of these two strains incubated with bacteriocin alone.

FIG. 2.

(A and B) Antimicrobial activity of AS-48 against B. cereus LWL1 in combination with sodium nitrite (A) and sodium lactate (B). Cultures were not treated with AS-48 (circles) or were treated with 5 μg of AS-48 per ml (squares) or 10μg of AS-48 per ml (triangles) alone (open symbols) or in combination with chemical preservatives (solid symbols). (C and D) Antimicrobial activity of AS-48 (5 μg/ml) against B. cereus LWL1 (C) and B. cereus CECT 131 (D) in the presence of no sodium chloride (•) or sodium chloride at a concentration of 3% (▪) or 5% (▴). Cultures were also treated with sodium chloride alone (○, no sodium chloride; □, 3% sodium chloride; ▵, 5% sodium chloride).

Production of enterotoxin by B. cereus LWL1 was studied in control cultures (OD₆₂₀, 0.1), as well as in the presence of bacteriocin concentrations ranging from 2.5 to 7.5 μg/ml. Samples of each culture were removed at different times (6, 10, 24, 48, and 72 h) after bacteriocin addition, centrifuged, filter sterilized, and tested to determine the reciprocal toxin titer by using a B. cereus reverse passive latex agglutination diarrheic enterotoxin detection kit (BCET-RPLA; Oxoid). Control cultures released high titers of enterotoxin after 6 h of cultivation, and the titers reached a maximum at 24 h (Fig. 3B). Bacterial growth and enterotoxin production were completely inhibited for at least 72 h by addition of AS-48 (7.5 μg/ml) at the beginning of the exponential growth phase (Fig. 3). The growth of cultures treated with subinhibitory concentrations of bacteriocin was markedly delayed, and the maximum titers of enterotoxin were 10-fold lower. A marked decrease in enterotoxin titers was also observed during prolonged incubation of bacteriocin-treated cultures (Fig. 3B). It seems unlikely that enterotoxin inactivation due to interaction with AS-48 occurred, because a single addition of AS-48 (5 μg/ml) to cultures after 24 h of incubation did not change the titers of previously produced enterotoxin compared to the titers in untreated control cultures during the subsequent 48 h (Fig. 3B).

FIG. 3.

Influence of AS-48 on growth (A) and enterotoxin production (B) of B. cereus LWL1. Early-exponential-phase cultures (time zero) were incubated with no bacteriocin AS-48 (○), 2.5 μg of AS-48 per ml (▾), 5 μg of AS-48 per ml (▴), or 7.5 μg of AS-48 per ml (▪). One culture was incubated without bacteriocin for 24 h before it was treated with 5 μg of AS-48 per ml (•).

One way to avoid toxin production in foods is to keep the size of the B. cereus population below 10⁷ cells/g (32) and to use inhibitory factors like acidic pH and glucose concentrations greater than 50 g/liter (36). Inhibition of enterotoxin production by AS-48 at a neutral pH should be useful for nonacidified foods.

To test if AS-48 had any effect on sporulation, bacteriocin was added at a subinhibitory concentration (5 μg/ml) to exponential-phase cultures of strain CECT 131 (OD₆₂₀, 0.1). Formation of heat-resistant spores both in controls and in treated cultures was monitored. Periodically, samples were heated at 80°C for 10 min to kill the vegetative cells (5), serially diluted, and plated on tryptic soy agar in triplicate to determine the number of spores. Sporulation of bacteriocin-treated cultures was delayed until the cultures recovered from bacteriocin treatment, and sporulation reached a maximum at 72 h (compared to 24 h in the controls).

Spores were obtained from B. cereus CECT 131 as described by Beuchat et al. (5) and were maintained at −80°C until they were used. The viability of dormant spores was not changed by incubation with AS-48 (50 μg/ml) for 3 h. Spores were induced to germinate with heat (70°C for 15 min) and then incubated for 1 h in an ice bath (24). To test the effect of AS-48 on release of dipicolinic acid (DPA) by germinating spores, two aliquots of a heat-shocked spore suspension (3.5 ml, 10⁹ spores/ml) in deionized water containing 50 mM l-alanine were dispensed into tubes containing 50 μl of distilled water (control) or AS-48 (final concentration 50 μg/ml). Both tubes were incubated at 37°C with shaking. At different times samples (500 μl) were taken from each tube and centrifuged in a microcentrifuge. The DPA in the supernatants was assayed by determining the DPA released, and the DPA in the spore pellet was assayed by determining the remaining DPA (22, 30); DPA obtained from Sigma was used as the standard. Identical sharp peaks of DPA release were observed in controls and in bacteriocin-treated spores within 5 min of induction of germination (data not shown), indicating that initiation of germination was not affected by AS-48.

Spores induced to germinate were incubated in prewarmed BHI at 37°C (10⁵ spores/ml). At different times germinating spores were incubated for 30 min with different bacteriocin concentrations and plated to determine viable cell counts (Fig. 4). Spores became sensitive to bacteriocin treatment 10 min after induction of germination, and high bacteriocin concentrations (25 to 50 μg/ml) caused a marked decrease in viable counts (Fig. 4). Sensitivity to AS-48 increased gradually during the course of germination. The greatest sensitivity to low bacteriocin concentrations was observed at 90 to 120 min, when vegetative growth started. Incubation of germinating spores with AS-48 at 5°C for 120 h had no effect on spore viability, but further incubation of the same mixture at 37°C resulted in a marked decrease in the viable count, like direct incubation of germinating spores with bacteriocin at this temperature. These results indicate that AS-48 was not inactivated during incubation at 5°C in the presence of spores. The activity of AS-48 that remained after prolonged incubation in the cold is an encouraging feature for use of this compound as an additive for foods and foodstuffs preserved at refrigeration temperatures to prevent temperature abuses during carriage or storage.

FIG. 4.

Effect of germination time on sensitivity of spores of B. cereus CECT 131 to bacteriocin AS-48. At different times during germination, spores were incubated for 30 min with different concentrations of AS-48.

The results obtained in this study are encouraging because bacteriocin AS-48 can reduce the sizes of large populations of B. cereus cells to levels below the B. cereus infective dose and can also inhibit toxin production. The reduction is accomplished at neutral pH values and is enhanced by some chemical preservatives, depending on the strain tested. These features, together with the ability of AS-48 to inhibit efficiently the proliferation of psychrotrophic strains and to decrease the viability of such strains at low temperatures, suggest that AS-48 may be a useful tool for preventing food poisoning by B. cereus. Therefore, in situ experiments should be carried out to determine the usefulness of AS-48 in foods.