INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Brochocin-C produced by Brochothrix campestris ATCC 43754 is a class IIb (two-peptide) bacteriocin that was originally reported by Siragusa and Cutter (22) and characterized by McCormick et al. (15). Brochocin-C has a broad activity spectrum comparable to that of nisin, and it is active against a broad range of gram-positive bacteria and spores of Clostridium and Bacillus spp. (11, 15). Nisin A is a well-characterized class I (lantibiotic) bacteriocin produced by Lactococcus lactis subsp. lactis. It has been widely accepted as a food preservative (6). Pediocin PA-1 produced by Pediococcus acidilactici PAC-1.0 is the most extensively studied class IIa bacteriocin (8).

Mechanistic studies of several lantibiotics and nonlantibiotics have revealed that their action occurs at the cytoplasmic membrane (7, 13, 19, 25). Nisin is not active against gram-negative bacteria, but liposomes of gram-negative bacteria (7) and sublethally heat-shocked gram-negative bacteria are inhibited by nisin A (5, 12). The outer membrane acts as a barrier to the action of nisin on the cytoplasmic membrane. Gram-negative bacteria treated with Tris-EDTA (24) and lipopolysaccharide (LPS) mutants of Salmonella enterica serovar Typhimurium (23) are sensitive to nisin A. Sublethally injured gram-negative bacteria are also susceptible to treatment with pediocin PA-1/AcH (12).

The object of this study was to determine whether the antibacterial activity of brochocin-C is comparable to that of nisin against gram-negative bacteria by determining if the outer membrane acts as a barrier to brochocin-C and by determining the effect of brochocin-C on the release of ATP from cells and spheroplasts of gram-negative bacteria.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains and growth media. Carnobacterium divergens NCFB 2855 (National Collection of Food Bacteria, Reading, United Kingdom; LV13), B. campestris ATCC 43754, and P. acidilactici PAC-1.0 (8) were grown in APT broth (All-Purpose Tween; Difco Laboratories, Detroit, Mich.) at 25°C. The gram-negative bacteria used in EDTA tests and Typhimurium LPS mutants were grown in brain heart infusion broth (Difco) at 37°C, except for Typhimurium strains LT2 and SL3600, which were grown in nutrient broth (Difco) and in proteose peptone-beef extract medium (18) for spheroplast preparation, respectively. All of the Salmonella LPS mutants were obtained from the Salmonella Genetic Stock Centre (University of Calgary, Calgary, Canada).

Partial purification of bacteriocins. Brochocin-C was partially purified (15) on Amberlite XAD-8 (BDH, Darmstadt, Germany), and the 60% ethanol eluent was concentrated to yield 25,600 arbitrary activity units (AU) of brochocin-C per ml. Pediocin PA-1/AcH was partially purified (10) by concentrating the 0.1% trifluoroacetic acid eluent from Sephadex G-50 chromatography (Pharmacia, Uppsala, Sweden) to yield 102,400 AU/ml. Purified nisin (Aplin and Barrett Ltd., Dorset, United Kingdom) was dissolved in 0.02 N HCl at 1 and 10 mg/ml, and stock solutions were stored at 70°C. All of the bacteriocin solutions were filter sterilized (Millex-GV filter; Millipore Corp., Bedford, Mass.). Activity (in AU per milliliter) was calculated as the reciprocal of greatest dilution that showed a clear zone of inhibition on a lawn of C. divergens LV13 by a spot-on-lawn assay (4).

Effect of brochocin-C on EDTA-treated gram-negative bacteria. Activity of the three bacteriocins was tested against strains of Escherichia coli and Salmonella spp. in the presence of 20 mM EDTA, as described by Stevens et al. (24). All of the bacteriocins were used at 3,200 AU/ml (for nisin, this was equivalent to 250 µg/ml). Viable bacterial counts of treated cells were determined before and after 30 min of treatment at 37°C. All experiments were repeated at least three times.

Effect of brochocin-C on Typhimurium LPS mutants. Typhimurium LT2 and its LPS mutants (Table 1) were grown at 37°C to an optical density at 600 nm of 0.15. The sensitivity of these strains to the bacteriocins was tested by a spot-on-lawn assay with brochocin-C at 800, 1,600, 3,200, 6,400, and 12,800 AU/ml, with nisin at 1,280, 12,800, and 64,000 AU/ml (0.1, 1, and 5 mg/ml, respectively), and with pediocin PA-1 at 204,800 AU/ml.

Effect of brochocin-C on spheroplasts of Typhimurium. Typhimurium LT2 and SL3600 (1% inoculum) were grown aerobically at 37°C. When the absorbance of LT2 at 600 nm reached 0.4, the culture was centrifuged at 3,000 × g and washed once and resuspended in sterile 10 mM Tris-HCl at pH 8.0. LT2 spheroplasts were prepared as described by Sambrook et al. (20), except that the lysozyme and EDTA concentrations were 8 mg/ml and 0.05 M, respectively. Spheroplasts were harvested after 15 min of incubation at 37°C. SL3600 spheroplasts were prepared as described by Osborn et al. (17). Spheroplasts were harvested by centrifugation at 3000 × g for 20 min and suspended in osmotically protected buffer (21). Formation of spheroplasts was monitored by phase-contrast microscopy. For use in experiments, spheroplasts were diluted to an absorbance at 600 nm of 0.5 to 0.6. The bacteriocins were added separately to a final concentration of 800 AU/ml. Suspensions with similar amounts of water were used as controls. Absorbance at 260 and 600 nm was monitored at selected time intervals during incubation at room temperature (23°C). To measure absorbance at 260 nm, samples were centrifuged at 3,000 × g for 20 min.

ATP leakage from bacteriocin-treated cells and spheroplasts. LT2 and SL3600 spheroplasts were suspended in osmotically protected buffer. Cells of LT2 and SL3600 were grown aerobically to an absorbance at 600 nm of 0.7. The cells from 50 ml of broth were harvested, washed once with 50 mM potassium phosphate buffer (pH 7.8), and resuspended in 2 ml of the same buffer. Cells and spheroplasts were stored on ice. Incubation mixtures consisting of 0.2 ml of cells or spheroplasts, 2.8 ml of 50 mM potassium phosphate buffer (pH 7.8), or osmotically protected buffer containing 0.5% glucose were held at room temperature. The bacteriocins were added to give final concentrations of 800 AU/ml. The protonophore CCCP (carbonyl cyanide m-chlorophenylhydrazone; Sigma Chemical Co., St. Louis, Mo.) was used at 40 µM as a control. Samples were taken at selected time intervals to determine ATP by bioluminescence assay (ATP bioluminescence assay kit; Sigma). The amount of bioluminescence emitted was integrated for 10 s and recorded in relative light units (luminometer model 1250; LKB Wallac, Bromma, Sweden). Extracellular ATP was determined in 100 µl of supernatant of samples centrifuged at 3000 × g for 20 min.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Inhibition of gram-negative bacteria. None of 29 strains of gram-negative bacteria was sensitive to brochocin-C, nisin, or pediocin PA-1. Treatment of E. coli ATCC 25922 and Salmonella enterica serovar Choleraesuis ATCC 10708 with EDTA and bacteriocins resulted in greater than a 2-log reduction in viable count in the presence of brochocin-C or nisin, but there was no reduction in the presence of pediocin PA-1 (Table 2). Viable counts of E. coli and serovar Choleraesuis treated with nisin and EDTA decreased by 4 and 2.1 logs, respectively, compared with 6.6- and 4.2-log reductions reported by Stevens et al. (24) for the same strains, even though a higher concentration of nisin was used in our experiments.

Inhibition of Typhimurium LPS mutants. Brochocin-C was inhibitory to only two of the three strains of Typhimurium mutants with an Re (heptoseless) chemotype (Table 3). This chemotype contains the least amount of LPS (only lipid A and 2-keto-3-deoxyoctonic acid). None of the LPS mutants was sensitive to pediocin PA-1. The Rc to Re chemotypes had different levels of sensitivity to nisin, and inhibition of Salmonella mutants increased with increasing concentrations. This was not the case with increasing concentrations of brochocin-C above 800 AU/ml.

Activity against spheroplasts. Addition of brochocin-C and nisin to spheroplasts of Typhimurium LT2 and its Re-type mutant SL3600 resulted in a decrease in absorbance at 600 nm (Fig. 1); however, when pediocin PA-1 was added, even at 3,200 AU/ml, no decrease in absorbance at 600 nm was observed during 170 min of exposure (data not shown). LT2 spheroplasts were more sensitive to brochocin-C than nisin, but the rate of action of brochocin-C against SL3600 spheroplasts was lower than that of nisin. The decrease in absorbance with brochocin-C and nisin treatment of LT2 spheroplasts was less than that observed in SL3600 spheroplasts. Brochocin-C and nisin only caused decreases in absorbance at 600 nm of 0.2 and 0.12, respectively. Furthermore, the reaction between the bacteriocins and spheroplasts was slow, requiring about 50 min for absorbance of nisin-treated LT2 spheroplasts to decrease 0.1 U, while absorbance of brochocin-treated LT2 spheroplasts decreased throughout the time of the experiment. During treatment for 170 min, there were 3.2- and 2.8-log reductions in the viability of SL3600 spheroplasts treated with brochocin-C and nisin, respectively, compared with 2- and 1.4-log reductions of LT2 spheroplasts.

View larger version (11K): [in this window] [in a new window]   FIG. 1.   Effect of brochocin-C and nisin (800 AU/ml, final concentration) on absorbance of spheroplasts of Typhimurium LT2 (A) and its LPS mutant SL3600 (B). Lysis was monitored at 600 nm. Symbols: , spheroplasts plus water; , spheroplasts plus nisin; , spheroplasts plus brochocin-C. Data are representative of three determinations.

ATP hydrolysis during treatment with bacteriocins. Absorbance at 260 nm of spheroplasts treated with brochocin-C and nisin increased, but not with pediocin PA-1 (data not shown). Total and extracellular ATP contents were determined on LT2 and SL3600 cells and spheroplasts treated with bacteriocins. The protonophore CCCP, which dissipates proton motive force, was used as a control. When spheroplasts were energized with 0.5% glucose, the total ATP concentration increased, reaching a maximum after 30 min, after which the bacteriocins were added (800 AU/ml). ATP hydrolysis was detected immediately after addition of brochocin-C or nisin to spheroplasts (Fig. 2). The ATP content of spheroplasts treated with pediocin PA-1 was similar to that of the negative control (data not shown). The total ATP dropped by more than 70% in 1 min for spheroplasts of LT2 treated with brochocin-C or nisin. Thereafter, only minor ATP hydrolysis occurred. Total ATP dropped to less than 10% immediately after addition of brochocin-C, nisin, or CCCP to SL3600 spheroplasts. External ATP was not detected in spheroplasts treated with brochocin-C (data not shown), while external ATP was 40 to 50% of total ATP in nisin-treated spheroplasts (Fig. 2). Similarly, the ATP concentration increased when LT2 and SL3600 cells were energized with glucose (Fig. 3). Upon addition of brochocin-C or nisin, the total ATP dropped within 1 min to 74 or 66% in LT2 cells and to 65 or 54% in SL3600 cells, respectively. Cells of LT2 and SL3600 treated with CCCP retained about 45% of their ATP and maintained these levels during the 70-min treatment period. LT2 cells treated with brochocin-C and nisin gradually replaced their ATP, and recovery was at 86% of ATP after 70 min of incubation in the presence of either bacteriocin. The external ATP of SL3600 cells treated with nisin was 30% of the total ATP (Fig. 3B), but external ATP was not detected in LT2 cells treated with nisin or in LT2 and SL3600 cells treated with brochocin-C (data not shown).

View larger version (12K): [in this window] [in a new window]   FIG. 2.   Effect of brochocin-C and nisin on total ATP levels in spheroplasts of Typhimurium LT2 (A) and its LPS mutant SL3600 (B) after 30 min of energizing with 0.5% glucose. The vertical arrow represents the time at which bacteriocins were added. The ATP levels are a percentage of the total ATP at 30 min. Symbols: , water; , brochocin-C; , nisin; *, CCCP; , nisin-induced ATP leakage. Data are means of three determinations.

View larger version (13K): [in this window] [in a new window]   FIG. 3.   Effect of brochocin-C and nisin on total levels of ATP in cells of Typhimurium LT2 (A) and its LPS mutant SL3600 (B) after 30 min of energizing with 0.5% glucose. The vertical arrow represents the time at which bacteriocins were added. ATP levels are a percentage of the total ATP at 30 min. Symbols: , water; , brochocin-C; , nisin; *, CCCP. In Fig. 3B, nisin-induced ATP leakage is also shown (). Data are means of three determinations.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Stevens et al. (24) showed that, in the presence of EDTA, nisin was active against cells of Salmonella and E. coli. We confirmed this with the same protocol by using two of the same target organisms, except that under our experimental conditions there was a smaller reduction in viable count. There was a comparable loss of viability of gram-negative bacteria when they were treated with brochocin-C and EDTA. Previously, we observed that brochocin-C might be toxic to E. coli when we used the general secretion pathway to produce this bacteriocin in this host (15). Access of bacteriocin to the cytoplasmic membrane is the key to activity of nisin and brochocin-C against gram-negative bacteria. Even though LPS mutants have the same chemotype, they do not necessarily have identical surface structures (26). This might explain why only Re strains SL3600 and SL1102 and not SA1377 are sensitive to brochocin-C. The ability of nisin and brochocin-C to penetrate the outer membrane of gram-negative cells most probably differs. The outer membrane can be made permeable to lysozyme by the use of a divalent ion chelator, such as EDTA, which loosens the structure of LPS. This leads to the disruption but not the complete removal of the outer membrane. Spheroplasts made with EDTA and lysozyme treatment contain some adherent outer membrane and entrapped murein (16).

Assuming that brochocin-C forms pores in the cytoplasmic membrane like other bacteriocins, pores formed by brochocin-C are smaller than those formed by nisin. ATP was released from SL3600 cells and spheroplasts, as well as LT2 spheroplasts treated with nisin, but not as a result of treatment with brochocin-C. Addition of nisin Z to Listeria monocytogenes resulted in hydrolysis and partial efflux of cellular ATP (3). Nisin forms transient multistate pores with a diameter ranging from 0.2 to 1.2 nm in black lipid membranes (1). This supports our observation that lysis of intracellular ATP by nisin is accompanied by substantial ATP leakage in spheroplasts and SL3600 cells. ATP leakage was not detected in LT2 cells, and intracellular decrease of ATP was less dramatic. Furthermore, cells of LT2 increase their rate of ATP production after treatment with nisin and brochocin-C (Fig. 3A). This might explain why brochocin-C and nisin do not affect viability of LT2. Spheroplasts of LT2 are more resistant to brochocin-C and nisin treatment than SL3600 spheroplasts (Fig. 2A), indicating that intact LPS provides protection to the cytoplasmic membrane.

Addition of brochocin-C to energized cells and spheroplasts resulted in a decrease in the intracellular ATP concentration, but no external ATP was detected. A similar observation was made with lactacin F on Enterococcus faecalis (2) and colicin A on E. coli (9). These authors proposed that ATP hydrolysis was caused by an efflux of inorganic phosphate resulting in a shift of the ATP hydrolysis equilibrium and/or the accelerated consumption of ATP to regenerate the decreased proton motive force. Dissipation of the proton motive force by CCCP reduced the intracellular ATP pool, indicating an enhanced use of ATP to regenerate the proton motive force. The effect of CCCP on the intracellular ATP pool was comparable in LT2 and SL3600 cells.

Two-peptide bacteriocins, such as lactacin F (2) and thermophilin 13 (14), form poration complexes in the cytoplasmic membrane. Thermophilin 13 is very similar in chemical structure to brochocin-C (15) and does not depend on membrane components from sensitive strains for its activity because it is active on liposomes (14). The results in this study might indicate that brochocin-C does not require a specific receptor for the bacteriocin to be active. This would explain why brochocin-C has a broad activity spectrum against gram-positive bacteria and it affects gram-negative bacteria when the outer membrane is impaired.