INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The ability of Listeria monocytogenes to resist environmental stresses has made this food-borne pathogen a major concern to the food industry. This pathogen is found on various foods and has been implicated in several large food-borne outbreaks worldwide (C. B. Dalton, C. Austin, J. Sobel, P. Hayes, B. Bibb, J. Mellen, and P. Griffin, Proc. 44th Annu. Epidemic Intelligence Serv. Conf., abstr. 19, 1995; 11). New preservation strategies have been developed to control the growth of L. monocytogenes in foods, including application of nisin (10, 15). Nisin is an antimicrobial peptide which kills L. monocytogenes as well as many other gram-positive bacteria. It acts on the cytoplasmic membrane of sensitive cells by forming transient pores which allow efflux of small hydrophilic compounds like ATP, ADP, monovalent cations, and amino acids (1, 25). This pore formation leads to dissipation of the membrane potential and ionic gradient across the membrane and subsequently results in the destruction of energy metabolism and cell death (3, 18).

The practical application of nisin as a food preservative can be compromised by the existence of L. monocytogenes strains which are naturally resistant to nisin (13, 17) and selection of nisin-resistant bacteria during exposure to progressively higher nisin concentrations (16, 17). However, the use of a multiple-hurdle system may reduce development of nisin resistance. Carbon dioxide inhibits growth of both gram-positive and gram-negative bacteria (6, 8). Our model studies and trials with cold-smoked salmon (20) have shown that a CO₂ atmosphere improves the antilisterial activity of nisin. Despite several publications on the effect of CO₂ on bacterial growth, the mechanism of its inhibitory activity still remains unclear.

The objective of this study was to gain mechanistic insight into the combined antilisterial effect of nisin and CO₂. Our data indicate that nisin and CO₂ atmosphere act synergistically on the cytoplasmic membrane of wild-type L. monocytogenes Scott A cells by enhancing membrane permeabilization. We also examined whether nisin and CO₂ have a synergistic antibacterial effect against a nisin-resistant (Nis^r) derivative of the Scott A strain. Alternate mechanisms of CO₂ action against wild-type Scott A cells, including inhibition of cell division, were investigated and excluded.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains and culture conditions. Wild-type L. monocytogenes Scott A and its Nis^r mutant L. monocytogenes ATCC 700302 (16) were grown in brain heart infusion (BHI) broth (Difco, Detroit, Mich.). The Nis^r strain was maintained on BHI agar (Difco) containing 25 µg of nisin per ml at 4°C. Lactococcus lactis strain NZ 9000 was transformed with recombinant plasmid pNZ8008 carrying the gusA gene fused to the nisA promoter (7) as described previously (14). The recombinant strain NZ 9000(pNZ8008) was grown at 30°C in M17 broth (Difco) supplemented with 0.5% (wt/vol) glucose (GM17) and chloramphenicol (10 µg/ml). This strain was used to determine nisin concentration (see below).

Chemicals. Nisin stock solutions were prepared from Nisaplin (a gift from Aplin and Barrett, Trowbridge, United Kingdom) for in vitro studies or pure nisin (Ambicin; a gift of AMBI Inc., Tarrytown, N.Y.) for in vivo studies. The stock solution was dissolved in nisin diluent (0.02 N HCl-0.75% NaCl, pH 5.3) and filter sterilized through a 0.45-µm-pore-size membrane filter (Acrodisc; Gelman Sciences, Ann Arbor, Mich.). When required, BHI was buffered (B-BHI) with 0.1 N phosphate buffer (K₂HPO₄-KH₂PO₄) to a pH of 6.2.

Growth of wild-type and Nis^r L. monocytogenes Scott A in the presence of nisin and in a CO₂ atmosphere. Wild-type and Nis^r cells of L. monocytogenes were grown at 30°C to an optical density at 600 nm (OD₆₀₀) of 0.5 to 0.6 (Shimadzu UV 160U; Tokyo, Japan). Wild-type cells were inoculated in B-BHI broth to give an initial cell concentration of approximately 5 × 10⁶ CFU/ml. Nisin was added at time zero at a concentration of 2.5 µg/ml. Cultures were incubated at 4°C with agitation (120 rpm) in air or in an atmosphere of 100% CO₂. To create a 100% CO₂ atmosphere, flasks were placed in anaerobic jars which had an external valve system (GasPak Jar System; BBL). Air was evacuated from the jars (0.9 bar), and the atmosphere composition was adjusted to 100% CO₂ by filling the anaerobic jars with CO₂ (gas purity was >99.8%; JWS Technologies). At appropriate intervals, CFU, MICs, and residual nisin concentrations were determined. CFU were determined by spiral plating (Model D; Spiral Biotech, Inc., Bethesda, Md.) of appropriate dilutions on BHI agar. After 24 to 30 h of incubation at 30°C, plates were counted with a laser bacterial colony counter (Model 500A; Spiral Biotech, Inc.).

Determination of nisin resistance. Nisin resistance in L. monocytogenes was measured by determining the MIC of nisin. L. monocytogenes cells were grown for 18 h at 30°C in BHI broth. Fresh BHI broth was inoculated with 10⁶ cells. The culture broth was inoculated into the wells of a microtiter plate (Corning Costar Corporation, Cambridge, Mass.), and nisin (12.5 µg/ml) was added in the first row of wells. Twofold dilutions of nisin in the culture were made, and the MIC was determined as the lowest nisin concentration which prevented L. monocytogenes growth after 24 h at 30°C.

Determination of nisin concentration by the -glucuronidase assay. The concentration of residual nisin in B-BHI broth was measured with the quantitative -glucuronidase assay. Lactococcus lactis NZ 9000(pNZ8008) cells were grown in M17 broth containing 0.5% (wt/vol) glucose (GM17) and 10 µg of chloramphenicol per ml at 30°C to an OD₆₀₀ of 0.5. Nisin was added as described by de Ruyter et al. (7). The specific -glucuronidase activity was measured in 96-well black plates with flat bottoms (Dynex Technologies), using an LS50B spectrometer (Perkin-Elmer) with excitation at 365 nm and emission at 455 nm. Nisin concentrations were derived from a dose-response standard curve (7).

Lipid extraction and fatty acid composition. L. monocytogenes Scott A cells were grown with aeration in B-BHI at 30 and 4°C, in the absence or presence of nisin (2.5 µg/ml), and in air or a 100% CO₂ atmosphere. Cells were harvested at mid-log phase (OD₆₀₀ of 0.6 to 0.8) and washed once with 0.1% peptone water (Difco), and lipids were extracted using the Bligh and Dyer (2) method, with modifications (25). Lipids were resuspended in chloroform-methanol (9:1) and stored at 20°C for up to 2 weeks before analysis for fatty acid composition (16).

Preparation of CF-loaded liposomes. Liposomes were made and loaded with carboxyfluorescein (CF) as previously described (25) and stored on ice for up to 4 h until use. The phospholipid concentration of liposomes was determined using the Bartlett assay as described by New (19).

CF release assay. The ability of nisin to cause CF efflux from liposomes derived from wild-type L. monocytogenes Scott A cells was studied in the presence of air and 100% CO₂ at 4 and 22°C. The effect of CO₂ at 22°C was not investigated since the solubility of the gas and consequently the antimicrobial effect are decreased at increasing temperatures (6). The release of CF from liposomes was determined as an increase in fluorescence intensity at 516 nm with excitation at 490 nm (F1T11 spectrofluorometer; Spex Industries, Metuchen, N.J.) as previously described (25). The assay buffer was changed to 0.1 N KHPO₄ due to its superior ability to stabilize pH in the presence of CO₂. The buffer's temperature was adjusted to the assay's temperature (4 or 22°C), and pH was adjusted to 6.2 with either 0.1 N K₂HPO₄ or 0.1 N KH₂PO₄ prior to each experiment. Precise temperature control was maintained by connecting a circulating temperature control bath (PolyScience, Niles, Ill.) to the cuvette holder in the spectrofluorometer. For the experiment in a 100% CO₂ atmosphere, the assay buffer (0.1 N KHPO₄, 4°C) was saturated with CO₂ by flushing 100% CO₂ directly into the buffer for approximately 5 min. The pH was then adjusted to 6.2 with either 0.1 N K₂HPO₄ or 0.1 N KH₂PO₄. During the assay, CO₂ was flushed into the cuvette chamber to saturate the space above the assay buffer with CO₂.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Influence of nisin and CO₂ on growth of Nis^r and wild-type L. monocytogenes Scott A. When grown in air at 4°C, Nis^r cells reached 8 × 10⁹ CFU/ml after 10 days (Fig. 1A). The growth patterns of the Nis^r strain in the presence and the absence of nisin were identical. Nisin (2.5 µg/ml) had no effect on Nis^r cells cultured in air or CO₂ atmospheres. A carbon dioxide atmosphere, on the other hand, extended the lag phase of the Nis^r culture for 6 days and decreased the final cell density by 2 log₁₀. Furthermore, carbon dioxide was more inhibitory to the growth of the Nis^r strain than to the growth of the wild-type strain (Fig. 1B).

View larger version (16K): [in this window] [in a new window]   FIG. 1.   Growth of L. monocytogenes Scott A Nisr (A) and wild-type (B) strains in B-BHI broth at 4°C. Closed symbols indicate results in air; open symbols indicate results in 100% CO2. Squares represent cultures in the absence of nisin, and circles represent cultures in the presence of nisin (2.5 µg/ml). Vertical bars represent the means and standard deviations from two independent experiments.

Effect of carbon dioxide on cell morphology of wild-type L. monocytogenes Scott A. The morphology of wild-type cells grown for 20 days in air and in the presence of 100% CO₂ atmosphere at 4°C is illustrated in Fig. 2. In the presence of 100% CO₂, cells elongated to two to five times their normal length.

View larger version (61K): [in this window] [in a new window]   FIG. 2.   Phase-contrast photomicrographs of L. monocytogenes Scott A (wild type) in the presence of air (A) and 100% CO2 (B).

Development of nisin resistance in wild-type cells. When the wild-type cells were grown in the presence of an initial concentration of 2.5 µg of nisin per ml, the MICs increased over time, both in air and in CO₂. For wild-type cells grown in air, the initial MIC was 3.1 µg/ml. This increased rapidly to 9.3 µg/ml after 3 days of incubation and to a maximum MIC of 12.5 µg/ml by day 10 (Fig. 3A). When grown in the presence of 100% CO₂, the MIC for the cells was stable at 3.1 µg/ml for 3 days, which was then increased gradually to 12.5 µg/ml by day 13. 

View larger version (19K): [in this window] [in a new window]   FIG. 3.   MICs of nisin (A) and residual concentrations of nisin (B) for L. monocytogenes wild-type cells grown at 4°C in the presence of air (black bars) or 100% CO2 atmosphere (open bars) when nisin (2.5 µg/ml) was added to the broth at time zero. Vertical bars represent the means and standard deviations from two independent experiments.

Influence of temperature and culture growth conditions on nisin-mediated CF efflux from liposomes. Temperature had a dramatic influence on the sensitivity of Listeria liposomes to nisin. CF-loaded liposomes were prepared using lipids extracted from cells cultured under three different conditions, 30°C in air, 4°C in air, and 4°C in a CO₂ atmosphere, in order to examine the effect of lipid composition. CF efflux was compared at two different temperatures to determine the role of membrane fluidity. For all three culture conditions examined, decreasing the assay temperature (and, thus, fluidity) markedly decreased the initial rates and overall extent of CF efflux and increased the time required for CF efflux to reach a plateau (Table 1). Liposomes derived from cells grown at 30°C in air were completely insensitive to nisin's action at 4°C. The negative CF efflux from liposomes of cells grown at 30°C in air when assayed at 4°C indicated that control liposomes (to which nisin diluent had been added) were slightly more leaky for CF than were liposomes treated with 2.5 µg of nisin per ml. The intrinsic (i.e., caused by diluent without nisin) leakiness of liposomes from cells grown in air at 30°C was 0.19%/min, 0.64% for liposomes from cells grown in air at 4°C, and 0.90%/min for cells grown at 4°C with a CO₂ atmosphere (data not shown).

View larger version (22K): [in this window] [in a new window]   FIG. 4.   Nisin-induced CF efflux from liposomes derived from wild-type L. monocytogenes Scott A grown in B-BHI broth at 30°C in air (a), 4°C in air (b), and 4°C in 100% CO2 (c). Nisin (2.5 µg/ml) was added at the time indicated by the arrow. The lipid concentration for cells preadapted at 30°C in air, 4°C in air, and 4°C in 100% CO2 was 2.5, 2.45, and 3.0 µM, respectively. The assay temperature was 22°C.

Influence of CO₂ atmosphere and growth conditions on nisin-mediated CF efflux at 4°C. When liposomes were exposed to a CO₂ atmosphere in the absence of nisin, there was a weak CF release over time, as previously noted. The overall CF efflux values (which had been corrected for baseline fluorescence) were only 0.77 to 1.9% after 600 S. Thus, the CO₂ atmosphere had some influence on the permeability of the liposomes. Interestingly, nisin-induced CF efflux was greatly enhanced in the presence of CO₂ atmosphere for all three types of liposomes examined (Fig. 5). Liposomes derived from cells grown at 30°C in air exhibited the greatest synergistic action between CO₂ and nisin. This was demonstrated by a 430% increase of the overall CF efflux at 1,200 s (Fig. 5A), albeit that the CF efflux levels were low. When assayed under air or a CO₂ atmosphere, liposomes derived from cells which had been grown at refrigeration temperature had higher CF efflux levels than did liposomes derived from cells grown at 30°C (Fig. 5B and C). Carbon dioxide atmosphere nonetheless increased CF efflux to higher levels. The overall CF efflux at 1,200 s increased by 186% (from 9.8 to 18.3%) and by 176% (from 30.1 to 53.2%) for liposomes derived from cells grown at 4°C in air (Fig. 5B) and at 4°C in 100% CO₂ (Fig. 5C), respectively. For all three types of liposomes, the combined effect of nisin and CO₂ on CF efflux was greater than the sum effects of CO₂ and nisin alone.

View larger version (22K): [in this window] [in a new window]   FIG. 5.   Influence of atmosphere composition on nisin-induced CF efflux from L. monocytogenes-derived liposomes. Liposomes were made from wild-type L. monocytogenes Scott A grown at 30°C in air (A), 4°C in air (B), and 4°C in 100% CO2 (C). Liposomes were exposed to air and 100% CO2, as listed next to the efflux curves. Nisin (2.5 µg/ml) was added at the time indicated by an arrow. The lipid concentration for cells preadapted to 30 and 4°C in air and 4°C in 100% CO2 was 2.5, 2.45, and 3.0 µM, respectively.

Membrane fatty acid composition of wild-type cells grown at 4°C. L. monocytogenes Scott A cells grown in the presence of 100% CO₂ had a greater proportion of short-chain fatty acids than did cells grown in air (Table 2). This was found both in the absence and in the presence of nisin. The primary response to a decrease in growth temperature from 30 to 4°C was a marked increase of monounsaturated fatty acids in the cell membrane composition (data not shown).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

These results demonstrated a quantifiable synergistic action of nisin and CO₂ on the cells (growth inhibition) and liposomes (membrane permeability) of wild-type L. monocytogenes Scott A. Carbon dioxide extended the lag phase and decreased the specific growth rate of both L. monocytogenes Scott A wild-type and Nis^r cells. CO₂ was more bacteriostatic against the Nis^r strain than was the wild-type strain. High CO₂ concentrations often extend the lag phase and inhibit growth (9). Results from this study verify that CO₂ acts similarly in L. monocytogenes cells. Wild-type cells grown in a 100% CO₂ atmosphere were two to five times longer than cells grown in air. Inhibition of cell division due to CO₂ has not, to our knowledge, previously been reported. However, CO₂ changes the cell morphology of Streptococcus mutans; at a high bicarbonate/K⁺ ratio, spherical cells are produced, and at a low bicarbonate/K⁺ ratio, the cells remain bacillary (23).

Lethality of nisin to L. monocytogenes cells and pore formation in liposomes were both enhanced in the presence of CO₂. Carbon dioxide reportedly acts like anaesthetic gases which expand hydrophobic regions of the membrane lipids. This perturbs membrane fluidity and could be expected to alter the function of a biological membrane (8). The increased CF efflux from liposomes derived from wild-type L. monocytogenes grown in the presence of CO₂ demonstrates that it destabilized the membrane. Moreover, adding CO₂ to the assay atmosphere enhanced the nisin-induced CF leakage. The correlation between our in vivo and in vitro data suggests that the synergistic action of CO₂ and nisin occurs at the cytoplasmic membrane.

In addition to increasing membrane permeability, carbon dioxide also caused the cells to modify their membrane fatty acid composition. The presence of CO₂ during growth of L. monocytogenes Scott A increased the proportion of short-chain fatty acids. The increase in short-chain fatty acids at the expense of long-chain fatty acids is suggestive of increased membrane fluidity (16). This in turn may enhance pore formation by nisin. Indeed, higher nisin-induced CF leakage was confirmed in the liposomes derived from cells grown in CO₂ than was confirmed in liposomes derived from cells grown in air.

The enhanced lethal action of nisin on cells grown in a CO₂ atmosphere could be attributed to a similar change in membrane fluidity. Carbon dioxide significantly reduces the MIC of nisin against L. monocytogenes cells grown at 10°C (20). The distinctive shift towards greater proportions of short-chain fatty acids, which is consistent with a more fluid membrane, would make cells adapted to a CO₂ atmosphere more susceptible to nisin. Resistance to nisin in L. monocytogenes is also accompanied by changes in lipid composition which are suggestive of a more rigid membrane (5, 16). The alterations in fatty acid profile may underlie the ability of CO₂ to delay the development of nisin resistance in wild-type cells.

Liposomes derived from cells grown at 4°C were dramatically more sensitive to nisin than were liposomes derived from cells grown at 30°C when CF efflux was measured at the same temperature. Extension of these in vitro findings to in vivo conditions would suggest that cells at low temperatures could be more sensitive to nisin, reducing the concern about high nisin tolerance developing in L. monocytogenes cells from refrigerated foods. Abee et al. (1) reported that nisin Z-induced K⁺ leakage from L. monocytogenes Scott A cells was greater in cells grown at 4°C than in cells grown at 30°C. This was explained by the increased proportion of unsaturated fatty acyl chains of the membrane lipids at 4°C, which helped to maintain an optimum membrane fluidity (1). Previously, we reported that the content of fatty acids with lower melting points is greater in L. monocytogenes cells grown at 10°C than in cells grown at 30°C (16). In agreement with these results, the present study found an increased proportion of monounsaturated fatty acids in the cell membrane at 4°C, compared to the proportion at 30°C. The alteration in fatty acid profile for cells cultured at the lower temperature, again suggestive of increased membrane fluidity, correlates with dramatically increased liposomal sensitivity to nisin.

Clearly, when the fatty acid composition is altered in response to culture conditions (i.e., CO₂ atmosphere and temperature) there have to have been changes in the nisin-induced CF efflux, presumably through modulation of membrane fluidity. The influence of assay temperature on the sensitivity of the liposomes to nisin further supports this membrane fluidity model. Temperature had a dramatic influence on pore formation of nisin. Decreasing the assay temperature of liposomes with the same lipid composition (i.e., derived from cells grown at a given condition) dramatically reduced nisin-mediated CF efflux. When the liposomes were switched to the lower temperature, a more-ordered fatty acid alignment may have resulted, making the membrane bilayer more rigid and thus more sensitive to nisin.

CO₂ atmosphere and nisin might be used in combination to control L. monocytogenes in refrigerated foods. The action of CO₂ on growing L. monocytogenes cells altered fatty acid composition, inhibited cell division, and delayed the development of nisin resistance. The synergistic action of nisin and CO₂ provides further support for a membrane fluidity model of interaction between nisin and the cytoplasmic membrane of a target cell. Several lines of evidence obtained from the present study suggest that membrane fluidity modulation in L. monocytogenes may underlie membrane poration by nisin. There is now a strong correlation between membrane fluidity and nisin sensitivity. The hypothesis that the observed changes in CF efflux (functional differences) are actually caused by changes in fluidity (difference in the physical state of membrane) is under investigation.