Inhibition of Bacillus anthracis Spore Outgrowth by Nisin

Abstract

The lantibiotic nisin has previously been reported to inhibit the outgrowth of spores from several Bacillus species. However, the mode of action of nisin responsible for outgrowth inhibition is poorly understood. By using B. anthracis Sterne 7702 as a model, nisin acted against spores with a 50% inhibitory concentration (IC₅₀) and an IC₉₀ of 0.57 μM and 0.90 μM, respectively. Viable B. anthracis organisms were not recoverable from cultures containing concentrations of nisin greater than the IC₉₀. These studies demonstrated that spores lose heat resistance and become hydrated in the presence of nisin, thereby ruling out a possible mechanism of inhibition in which nisin acts to block germination initiation. Rather, germination initiation is requisite for the action of nisin. This study also revealed that nisin rapidly and irreversibly inhibits growth by preventing the establishment of oxidative metabolism and the membrane potential in germinating spores. On the other hand, nisin had no detectable effects on the typical changes associated with the dissolution of the outer spore structures (e.g., the spore coats, cortex, and exosporium). Thus, the action of nisin results in the uncoupling of two critical sequences of events necessary for the outgrowth of spores: the establishment of metabolism and the shedding of the external spore structures.

Lantibiotics are methyllanthionine-containing cationic antimicrobial peptides produced by several gram-positive bacteria (11). Nisin is a 34-amino-acid peptide produced by Lactococcus lactis subsp. lactis (ATCC 11454), which has emerged as an important prototype for the study of the novel antibacterial properties and structure-activity relationships characteristic of the lantibiotics (5, 33). Like all lantibiotics, nisin is ribosomally translated and is then posttranslationally modified to generate three noncyclic nonproteogenic amino acids, dehydroalanine, and dehydrobutyrine and five lanthionine or methyllanthionine thioether rings (11).

The utility of nisin derives from its capacity to act upon gram-positive bacteria by two entirely different mechanisms (15, 46). Nisin forms pores in lipid membranes (46), but it also functions as a transglycosylase inhibitor that disrupts cell wall biosynthesis via lipid II binding and mislocalization (21, 55). Because it functions as a “two-edged sword,” microbes have been relatively refractory to the emergence of resistance to nisin, despite its widespread and persistent use as a preservative in the food industry (15, 46).

An additional and poorly understood activity of nisin is its capacity to prevent the outgrowth of spores from several gram-positive bacteria, including several Bacillus species (9, 10, 40, 42). To date, nisin inhibition of Bacillus spore outgrowth has been documented by various methods, including the spectrophotometric measurement of liquid culture turbidity (3), the enumeration of CFU (4, 14, 32, 35, 43), well diffusion assays on solid agar (14, 39), and microscopic observations (41). Although these approaches are useful, they have provided few details about nisin's mode of action against Bacillus spores. Currently, it has not been experimentally established whether nisin inhibits spore outgrowth by preventing germination initiation or, alternatively, preventing a step downstream of germination initiation. Additionally, the requirement for germination for the action of nisin has not been addressed. Finally, it is not clear whether or not the action of nisin requires actively growing organisms, analogous to many other antibiotics.

To address these issues, the effects of nisin on Bacillus spores and their development into replicating bacilli were evaluated by using spores from Bacillus anthracis Sterne 7702 as a model. The results from these studies indicate that nisin does not inhibit germination initiation; instead, germination is required for irreversible inhibition. Nisin acted rapidly upon germinating spores to prevent the establishment of oxidative metabolism or the membrane potential, possibly by a mechanism involving the disruption of membrane integrity. Nisin did not inhibit the removal of the outer spore structures (e.g., the exosporium, cortex, and spore coat). Collectively, these data suggest that nisin acts upon spores immediately after the initiation of germination and effectively blocks the capacity of B. anthracis to proliferate and produce virulence factors.

MATERIALS AND METHODS

Spore preparations.

Spores were prepared from B. anthracis Sterne 7702, as described previously (49). The enumeration of spores or bacilli was performed with a Petroff-Hauser hemocytometer under a light microscope at ×400 magnification (Nikon Alphaphot YS, Mellville, NY). A typical spore preparation yielded 10 ml of spores at a concentration of 2.0 × 10⁹ spores/ml.

Nisin purification.

A sample of 500 mg nisaplin (50% denatured milk proteins, 2.5% nisin, 47.5% sodium chloride) was suspended in 30% acetonitrile (Sigma, St. Louis, MO) with 0.1% trifluoroacetic acid (10 ml; Sigma). The suspension was sonicated for 20 min, followed by centrifugation at 1,500 × g for 10 min to remove all insoluble material. Reverse-phase high performance chromatography (Waters, Milford, MA) was performed with a PrePack C₄ semipreparative column (diameter, 25 mm; length, 100 mm; Waters) with a gradient of 0 to 100% acetonitrile. Under these conditions, nisin had a retention time of 28 min. Acetonitrile and trifluoroacetic acid were removed from fractions containing nisin by rotary evaporation, followed by lyophilization to remove the water. Prior to use, lyophilized nisin was weighed on an analytical balance and was dissolved in 0.1 M 3-(N-morpholino)propanesulfonic acid (MOPS; pH 6.8) to yield the desired concentration. The identity of purified nisin was confirmed by matrix-assisted laser desorption ionization-time of flight mass spectrometry (General Electric, NY). As an additional quality control measure, purified nisin was evaluated for its inhibitory activity against Lactococcus lactis 117 (ATTC 15577) cells grown in GM17 broth (3.7% M17 medium, 0.5% dextrose; BD Biosciences) at 30°C. Purified nisin inhibited L. lactis 117 with a 50% inhibitory concentration (IC₅₀) of 0.0021 μM, in excellent agreement with the findings of previous studies (6, 27, 28), indicating that the purification protocol yielded nisin with the expected biological activity.

Culture of B. anthracis spores.

B. anthracis Sterne 7702 spores at a concentration of 4.0 × 10⁶ spores/ml, unless indicated otherwise, were incubated in brain heart infusion (BHI) medium supplemented with nisin (0.1, 1, 10, and 100 μM), ciprofloxacin (0.01, 0.1, 1, and 10 μM), or 0.1 M MOPS (pH 6.8; Sigma) as a mock control. In these studies, the changes in germinating spores caused by nisin were compared to those induced by ciprofloxacin, an antibiotic recommended for use for the treatment of anthrax. The published MIC of ciprofloxacin against B. anthracis is 0.193 μM (34), which was consistent with the results obtained in preliminary experiments (data not shown), thus providing a basis for the range of ciprofloxacin concentrations used in these studies. Each of the ciprofloxacin studies was repeated twice with independent preparations of spores. For nongerminating conditions, 0.1 M MOPS (pH 6.8) was substituted for BHI medium. All incubations were performed at 37°C under aeration (180 rpm on a rotary shaker [Thermo Fisher Scientific Inc., Waltham, MA] or as indicated otherwise) and ambient CO₂ (e.g., 0.03% CO₂). In pilot experiments, spores were incubated in alternative media, which were Luria-Bertani (LB; 10 g/liter Bacto tryptone, 5 g/liter NaCl, 5 g/liter Bacto yeast extract; BD Diagnostics), RPMI 1640 medium (ATCC) containing fetal bovine serum (FBS; 10%; JRH Biosciences, Lenexa, KA), minimal essential medium (MEM; JRH Biosciences) containing FBS (10%), or Dulbecco's MEM (DMEM; JRH Biosciences) containing FBS (10%).

Determination of IC₅₀s and IC₉₀s of nisin against endospores.

B. anthracis endospores at a final concentration of 4.4 × 10⁴, 4.4 × 10⁵, 4.4 × 10⁶, or 4.4 × 10⁷ spores/ml were incubated in BHI medium supplemented with various concentrations of nisin (0.05 μM to 100 μM) or 0.1 M MOPS pH 6.8 (as a negative control). The IC₅₀s and IC₉₀s were derived from plots of the optical density at 600 nm (OD₆₀₀) at 16 h versus the nisin concentration and are the concentrations of nisin that inhibited B. anthracis growth in BHI medium by 50% and 90%, respectively.

CFU quantification.

Spores were serially diluted and plated on agar plates containing LB medium (10 g/liter Bacto tryptone, 5 g/liter NaCl, 5 g/liter Bacto yeast extract, 15 g/liter Bacto agar; BD Biosciences). After 12 to 18 h at 37°C, the B. anthracis colonies were counted, and the numbers of CFU/ml were calculated from those counts.

Spore hydration.

The hydration of spores was determined by measuring the loss of spore refractility at 600 nm by using a Synergy 2 plate reader (BioTek Instruments, Inc., Winooski, VT). B. anthracis spores were incubated, as described under “Culture of B. anthracis spores,” except that a 96-well plate was used and the plate was shaken for 15 s prior to each read. The data are presented as a percentage of the OD₆₀₀ at each time point relative to the OD₆₀₀ of the spore suspensions at the beginning of the experiment (time zero).

Heat resistance.

Spores were diluted into 0.1 M MOPS (pH 6.8) containing d-alanine and d-histidine (both at 10 mM; Sigma), to prevent the further germination initiation of dormant spores, and identical aliquots were incubated at either 65°C or on ice for 30 min. Viable B. anthracis organisms were quantified by plating serial dilutions and enumerating the CFU. The percentage of heat-resistant spores was calculated by dividing the numbers of CFU recovered from the samples heated at 65°C by the numbers of CFU recovered from the samples incubated on ice.

DIC microscopy.

At the indicated times, samples were removed from the B. anthracis cultures and fixed by incubation in 3% formaldehyde (Sigma) for 30 min at 37°C, followed by the mounting of samples on glass slides in 20% glycerol (Sigma). Differential interference contrast (DIC) microscopy images were collected with an Applied Precision assembled DeltaVision epifluorescence microscope containing an Olympus Plan Apo ×100 oil objective with a numerical aperture of 1.42 and a working distance of 0.15 mm, and the images were processed with the SoftWoRX (Issaquah, WA) Explorer Suite program.

Immunoblot analysis.

At the indicated times, samples removed from B. anthracis cultures grown in the presence of 0.2% (wt/vol) bicarbonate at 37°C under 5% CO₂ were centrifuged for 10 min at 21,000 × g. The culture supernatants were denatured by the addition of an equal volume of 2× sodium dodecyl sulfate (SDS) sample buffer (4% SDS, 100 mM Tris, 0.4 mg bromophenol blue/ml, 0.2 M dithiothreitol, 20% glycerol). The samples were boiled for 5 min and were resolved by SDS-polyacrylamide gel electrophoresis (10% acrylamide). The contents of the gels were electrotransferred to nitrocellulose membranes (Pierce, Rockford, IL). The membranes were probed for the presence of protective antigen (PA) and lethal factor (LF) by utilizing anti-PA (QED Bioscience Inc., San Diego, CA) and anti-LF (QED Bioscience Inc.) mouse monoclonal antibodies, respectively. Goat horseradish peroxidase-conjugated anti-mouse immunoglobulin G (Abcam Inc., Cambridge, MA) was used as the secondary antibody, and cross-reacting material was visualized after the blots were exposed to X-ray film (Denville Scientific Inc., Metuchen, NJ) in the presence of the enhanced chemiluminescence immunoblotting reagent (Pierce, Rockford, IL). For the experiments for the investigation of an association between the presence of PA or LF with spores, spore homogenates were prepared by vortexing spore suspensions 10 times with 0.1-mm-diameter glass beads for 30 s.

Oxidative metabolism.

Samples from each culture were diluted into 0.1 M MOPS (pH 6.8) containing d-alanine and d-histidine (both at 10 mM; Sigma), to prevent the further germination initiation of dormant spores. Each sample was then incubated with 3-(4,5-di-methylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (tetrazolium; 5 mg/ml) for 30 min at 37°C. The conversion of tetrazolium to formazan was measured at 570 nm with a Synergy 2 plate reader (12).

Membrane potential.

The B. anthracis spores were incubated as described under “Culture of B. anthracis spores,” except for the presence of a fluorescent membrane potential-sensitive dye, 3-3′diethyloxacarbocyanine iodide (DiOC₂; 300 nM; Invitrogen, Carlsbad, CA) (29). At the indicated times, the membrane potential was assessed by measuring the increase in B. anthracis-associated DiOC₂ fluorescence by flow cytometry (EPICS XL-MCL flow cytometer; Beckman Coulter, Fullerton, CA), with excitation at 488 nm with an argon laser and measurement of the fluorescence emission through a band-pass filter at 525/20 nm. At least 10,000 events were detected for each sample, and the data were analyzed by using the FCS Express 3.00.0311 V Lite Standalone software. The data were plotted as the geometric mean of the fluorescence intensity (MFI).

Membrane integrity.

Membrane integrity was evaluated by measuring the uptake of propidium iodide (PI) (19, 50). Samples from each culture were incubated with PI (60 μM; Molecular Probes Inc., Leiden, The Netherlands) in an ice bath for 10 min (1). B. anthracis-associated fluorescence was measured by flow cytometry as described above under “Membrane potential,” except that the fluorescence emission was measured with a band-pass filter at 675/20 nm.

Quantification of DPA.

The release of dipicolinic acid (DPA; 2,6-pyridinedicarboxylic acid) was monitored by measuring the fluorescence resonance energy transfer between DPA and terbium (25, 45). B. anthracis spores were incubated in a manner similar to that described above under “Culture of B. anthracis spores,” except for the presence of TbCl₃ (200 μM; Sigma). The DPA-terbium complex was excited at 280 nm, and emission was monitored at 546 nm with a Synergy 2 plate reader.

TEM.

The B. anthracis organisms from each culture were concentrated by centrifugation (21,000 × g for 30 min), the pellets were resuspended in Karnovsky's fixative (26), and samples were prepared for transmission electron microscopy (TEM) analysis, as described previously (56). Images were collected with a CMI Hitachi H600 transmission electron microscope (Tokyo, Japan) in the University of Illinois College of Veterinary Medicine Microscopy Facility.

Statistics.

Error bars represent standard deviations. P values were calculated by Student's t test by using a paired, one-tailed distribution. A P value of <0.05 indicates statistical significance.

RESULTS

Growth inhibition by nisin.

Previous studies reported that nisin prevented the growth of Bacillus spores derived from several different species (4, 14, 32, 35, 43). To evaluate the action of nisin against spores responsible for this inhibitory activity, spores were prepared from B. anthracis Sterne 7702, which is a strain commonly employed as a model for investigation of the early stages of anthrax disease (24, 47). To validate earlier results, B. anthracis spores were incubated in BHI medium, which has been used to induce the germination of B. anthracis spores (17), supplemented with either nisin (0.05 μM to 100 μM) or buffer control (0.1 M MOPS, pH 6.8). Pilot experiments confirmed that 0.1 M MOPS (pH 6.8) alone does not induce spore germination (data not shown). In BHI medium inoculated with 4.4 × 10⁶ spores/ml, nisin inhibited B. anthracis growth, with IC₅₀s and IC₉₀s of 0.57 μM and 0.90 μM, respectively (Table 1). The inhibitory activity of nisin against spores was not strictly dependent on BHI medium, as nisin inhibited B. anthracis growth to the same degree in LB medium or in MEM, DMEM, or RPMI 1640, each supplemented with 10% FBS (data not shown). Relative to the numbers of CFU in cultures at the initial time point, approximately 10,000-fold more CFU was recovered at 10 h from cultures supplemented with either 0.1 μM nisin or 0.1 M MOPS (pH 6.8). In contrast, no detectable CFU was recovered from 10-h cultures that had been supplemented with 1, 10, or 100 μM nisin (data not shown).

TABLE 1.

IC₅₀ and IC₉₀ of nisin against B. anthracis spores^a

No. of spores/mlb	IC50 (μM)c	IC90 (μM)d
4.4 × 104	0.17 ± 0.01	0.41 ± 0.02
4.4 × 105	0.19 ± 0.01	0.44 ± 0.01
4.4 × 106	0.57 ± 0.03	0.90 ± 0.01
4.4 × 107	0.63 ± 0.06	0.98 ± 0.01

^aThree independent experiments were performed in triplicate with different spore preparations and nisin purifications. The values are reported as the averages of three experiments.

^bSpores were freshly prepared from B. anthracis Sterne 7702.

^cDefined as the nisin concentration that inhibits the growth of cultures of B. anthracis spores by 50% at 16 h.

^dDefined as the nisin concentration that inhibits the growth of cultures of B. anthracis spores by 90% at 16 h.

On the basis of the results of these experiments, subsequent studies were conducted with 4.0 × 10⁶ spores/ml because this concentration of spores was sufficient to generate detectable readouts for each assay and yielded IC₅₀s and IC₉₀s similar to those calculated with a higher spore concentration (4.4 × 10⁷ spores/ml; Table 1) and at the same time allowed several full sets of experiments to be conducted from each spore preparation.

Nisin does not inhibit germination initiation.

The inability to recover CFU from cultures of B. anthracis spores supplemented with nisin could be due to the irreversible inhibition of germination initiation. To evaluate this possibility, germination initiation was first monitored by measuring the characteristic loss of spore refractility that accompanies hydration of the spore structure, as indicated by a decrease in the OD₆₀₀ (30, 53). These experiments revealed a loss in refractility of >65% by 10 min in both the presence and the absence of nisin (Fig. 1A), indicating that nisin did not detectably alter hydration of the spores following germination initiation. Similarly, the loss of spore refractility was not inhibited in the presence of even the highest tested concentration of ciprofloxacin (10 μM), a DNA gyrase inhibitor that acts upon bacteria by a mechanism fundamentally different from that of nisin. Incubation of spores with nisin or ciprofloxacin alone (in the absence of known germinants) did not result in a loss of spore refractility (data not shown). These results are consistent with the notion that neither nisin nor ciprofloxacin inhibits B. anthracis growth by blocking germination initiation.

FIG. 1.

Nisin does not alter germination initiation. (A) The data are expressed as the percentage of the OD₆₀₀ at time zero and 10 min relative to that of each culture at time zero. Shown is the mean of a single experiment conducted in triplicate as a representative of three independent experiments. Error bars indicate standard deviations. In all cases, the differences between spore refractility at 10 min relative to that at 0 min were statistically significant (P < 0.05). (B) At 0 and 5 min, samples were analyzed for heat resistance, as described under Materials and Methods. The data are expressed as the means of three experiments. Error bars indicate standard deviations. In all cases, the differences between the percentage of heat-resistant spores at 5 min relative to that at 0 min were statistically significant (P < 0.05).

A second hallmark of germination initiation is the rapid loss of spore resistance to heat (30, 53). In both the presence and the absence of nisin, spores demonstrated a >80% loss in heat resistance by 5 min (Fig. 1B), providing additional evidence that germination initiation was not altered in the presence of nisin. When spores were incubated in 0.1 M MOPS (pH 6.8) supplemented with nisin, no loss of heat resistance was observed (data not shown), again confirming that nisin does not induce germination. Taken together, these results indicate that the loss of recoverable CFU from cultures of spores supplemented with nisin was not due to the inhibition of germination initiation, thereby ruling out the possibility that such a mechanism underlies the inhibitory action of nisin against spores. Instead, these data would seem to indicate that the loss of recoverable CFU from cultures of spores supplemented with nisin may be due to nisin-mediated killing of the germinated spores.

Germination initiation is required for the inhibitory action of nisin.

Whether germination initiation is necessary for nisin to act against spores was investigated next. Spores were incubated in BHI medium, in BHI medium supplemented with 10 μM nisin, and in 0.1 M MOPS supplemented with 10 μM nisin. After 1 h, the spores were washed to lower the concentration of nisin in solution to approximately 1 nM, which is well below the IC₅₀. After the spores were washed, they were introduced into fresh BHI medium. As expected, spores that had been preincubated with nisin under germinating conditions did not grow when they were introduced into fresh BHI medium (Fig. 2). In contrast, spores preincubated with nisin in the absence of germinant demonstrated robust growth in fresh BHI medium. These data indicate that germination initiation is requisite for the inhibitory activity of nisin against spores.

FIG. 2.

Germination is required for the inhibitory action of nisin. The data are expressed as the mean of a single experiment conducted in triplicate and are representative of those from two independent experiments. Error bars indicate standard deviations. In all cases, the differences between the OD₆₀₀ at 18 h relative to that at 0 h were statistically significant (P < 0.05).

To establish the point during the germination process at which nisin-mediated inhibition becomes irreversible, spores were preincubated in BHI medium (to induce germination) supplemented with either nisin (10 μM) or 0.1 M MOPS. At various times, samples from each culture were washed extensively to lower the concentration of nisin in the spore suspensions to levels well below the IC₉₀ and were introduced into fresh BHI medium. These experiments revealed that the exposure of B. anthracis spores to nisin for as few as 5 min under germinating conditions completely blocked the growth of the germinated spores in fresh medium lacking nisin (Fig. 3). These results suggest that the inhibitory action of nisin against spores becomes irreversible soon (<5 min) after germination is initiated.

FIG. 3.

The inhibitory action of nisin is irreversible. The data are expressed as the mean of a single experiment conducted in triplicate and are representative of those from two independent experiments. Error bars indicate standard deviations. In each case in which the spores were exposed to nisin, the increase in the OD₆₀₀ at 18 h relative to that at 0 h was not statistically significant (P > 0.05).

Nisin prevents spore development into vegetative bacilli.

To obtain additional insights into the stage of germination at which nisin arrests the development of spores into replicating, vegetative bacilli, the extended growth of cultures was monitored in the presence or the absence of nisin. As expected, the OD₆₀₀ of each of the samples decreased initially, reflecting the rapid hydration of spores that characteristically follows germination initiation (Fig. 4A). After approximately 50 min, cultures supplemented with 0.1 μM nisin demonstrated clear bacterial growth, albeit at a lower rate than cultures lacking nisin (Fig. 4A). In contrast, there was no evidence of growth in cultures supplemented with higher concentrations of nisin either at 180 min (Fig. 4A) or at extended time points (12 h; data not shown). Examination of samples removed from these cultures at 5 and 10 h by DIC microscopy revealed characteristic chains of vegetative bacilli from cultures supplemented with either 0.1 μM nisin or the buffer control (Fig. 4B), whereas no bacilli were present within cultures supplemented with 1, 10, or 100 μM nisin (Fig. 4B). These results indicate that at concentrations greater than the IC₉₀, nisin prevents the development of germinated spores into vegetative bacilli. By comparison, germinated spores developed into vegetative bacilli by 5 h in the presence of 1 μM ciprofloxacin (data not shown), which is a concentration sufficient to completely inhibit the proliferation of vegetative bacilli.

FIG. 4.

B. anthracis spores do not develop into vegetative bacilli in the presence of nisin. (A) The data are expressed as the percentage of OD₆₀₀ at each time point relative to the OD₆₀₀ of each culture at time zero, which was the control in these experiments. The data are expressed as the means of a single experiment conducted in triplicate and are representative of those from three independent experiments. Error bars indicate standard deviations. (B) At time zero and 5 and 10 h, samples were removed and visualized by DIC microscopy. For each panel, a single spore is shown for clarity but is representative of all other B. anthracis spores within that sample. Bars, 6.5 μm. The data are representative of those from three independent experiments.

Spores incubated with nisin do not produce lethal toxin.

Although nisin prevents the development of spores into vegetative bacilli, the extent to which the action of nisin impairs additional events associated with spore germination was not clear. For B. anthracis, an important consideration is whether or not virulence factors, such as lethal toxin (LT) (51), are released prior to nisin-mediated killing of germinated spores. Spores were incubated under conditions that are known to induce LT production (23). In the absence of nisin, the two components of the bipartite LT, LF and PA, were both readily detected by immunoblot analysis (Fig. 5A). In contrast, neither LF nor PA was detected within culture supernatants prepared from B. anthracis cultures supplemented with 10 μM nisin. By comparing the intensity of the cross-reacting material in twofold serially diluted cultured supernatants, the amounts of PA and LF were determined to be reduced at least 32- and 16-fold, respectively, in cultures supplemented with nisin compared to the amounts in cultures supplemented with the buffer control (0.1 M MOPS, pH 6.8; data not shown). Additional experiments showed that neither LF nor PA was detected within homogenates of B. anthracis spores incubated for 1 h in BHI medium supplemented with nisin (data not shown), ruling out the possibility that LF or PA was present but spore associated. Finally, pilot experiments indicated that nisin did not cause either LF or PA to precipitate out of solution (data not shown), ruling out another possible explanation for the absence of these proteins in B. anthracis cultures supplemented with nisin.

FIG. 5.

Nisin prevents B. anthracis spores from becoming metabolically active. (A) At 0, 7, and 10 h, culture supernatants were evaluated for the presence of LF and PA by immunoblot analysis. The samples in each lane were normalized for the volumes of the culture supernatants. The data are from a single experiment and are representative of data collected in three independent experiments. (B) At the indicated times, aliquots were removed from the cultures and were evaluated for oxidative metabolism by measuring spectrophotometrically the production of formazan at 570 nm, as described under Materials and Methods. (C) At time zero (i.e., prior to the addition of nisin) and 30 min, aliquots were removed from the cultures and evaluated for the membrane potential by measuring the DiOC₂-associated B. anthracis fluorescence by flow cytometry. The data are plotted as the MFI. (B and C) Means of the data from a single experiment conducted in triplicate. The data are representative of those from three independent experiments. Error bars indicate standard deviations. For each sample in panel C incubated in BHI medium, the difference between the membrane potential at 30 min in the presence and the absence of nisin was statistically significant (P < 0.05).

The action of nisin prevents spores from becoming metabolically active.

Dormant B. anthracis spores are metabolically inactive (48). One potential explanation for the lack of detectable LF or PA in B. anthracis cultures supplemented with nisin is the inability of germinating spores to establish an active metabolism. To evaluate this hypothesis, the cellular production of NAD(P)H was monitored as a measure of oxidative metabolism by determining the reduction of tetrazolium to formazan in an NAD(P)H-dependent manner (12). In the absence of nisin, the robust production of formazan was detected, beginning at 5 to 10 min after the initiation of germination, and the levels of formazan generated continued to increase during the course of the experiment (3 h) (Fig. 5B). In the presence of 1, 10, or 100 μM nisin, small but detectable levels of formazan production were detected within 5 to 10 min after the initiation of germination, but formazan production did not continue to increase after this time. Formazan production was detected in the presence of 0.1 μM nisin, albeit at a lower rate than in the absence of nisin. Formazan production was inhibited in the presence of 1 and 10 μM ciprofloxacin (data not shown) but occurred in the presence of 0.01 and 0.1 μM ciprofloxacin, albeit to a lesser extent (approximately 65% and 40%, respectively, compared to the amount of formazan produced in the absence of antibiotic).

Because oxidative metabolism is linked to the establishment of an electrochemical gradient across the cytoplasmic membrane, the effects of nisin on the establishment of a membrane potential within germinating spores were evaluated. In the presence of germinant, B. anthracis demonstrated significantly stronger staining with the membrane potential-sensitive dye DiOC₂ (29) than in the absence of germinant, indicating the establishment of a membrane potential by 30 min subsequent to germination initiation (Fig. 5C). In contrast, spores incubated in the presence of nisin demonstrated significantly less DiOC₂ staining at 30 min (Fig. 5C), which indicated that at this early time point, nisin interfered with the establishment of a membrane potential in germinating spores. By 5 and 10 h after germination initiation, spores incubated in the presence of 0.1 μM nisin demonstrated DiOC₂ staining similar to that of spores in the absence of nisin (94.9% MFI of spores in the absence of nisin; data not shown), indicating that these spores recovered and ultimately developed a membrane potential, albeit at a lower rate. Spores incubated in the presence of higher concentrations of nisin (1, 10, and 100 μM) did not demonstrate increased DiOC₂ staining at later time points (5 or 10 h; data not shown). Notably, in the presence of ciprofloxacin (0.01, 0.1, 1, and 10 μM), germinating spores displayed DiOC₂ staining comparable to that in the absence of antibiotic (approximately 80 to 100%; data not shown), indicating that, in contrast to nisin, ciprofloxacin did not prevent the establishment of a membrane potential. Taken together, these studies suggest that nisin acts upon spores immediately after the initiation of germination and that at concentrations nonpermissive for spore outgrowth (as demonstrated in Fig. 4) nisin prevents B. anthracis from becoming metabolically active.

Effects of nisin action on membrane integrity.

The absence of oxidative metabolism in germinating spores in the presence of nisin could be due to a loss of membrane integrity (19, 50). To explore this possibility, germinating spores were evaluated for increases in membrane permeability by measuring the uptake of PI by flow cytometry. These experiments revealed that by 30 min, nisin induced 2-, 6-, 13-, and 56-fold increases in the amount of PI taken up by spores incubated with 0.1, 1, 10, and 100 μM nisin, respectively, relative to the amount taken up by spores incubated in the absence of nisin (Fig. 6). The results of thse experiments suggest that within germinating B. anthracis spores, nisin induces a dose-dependent disruption of membrane integrity. In contrast, germinating spores exhibited only a modest increase in PI uptake (less than twofold; data not shown) in the presence of ciprofloxacin (0.01, 0.1, 1, or 10 μM), further supporting the idea that nisin and ciprofloxacin inhibit the outgrowth of B. anthracis spores by fundamentally different mechanisms.

FIG. 6.

Effects of nisin on B. anthracis membrane integrity. At the indicated times, aliquots were removed from the cultures and evaluated for PI uptake, as described under Materials and Methods. The data were plotted as the geometric MFI. The means of the data from a single experiment conducted in triplicate are presented. The data are representative of those from three independent experiments. Error bars indicate standard deviations. In all cases, the differences in PI uptake in samples containing nisin at 30 and 60 min relative to that at 0 min were statistically significant (P < 0.05).

Nisin does not prevent spore remodeling during germination.

The inhibition of spore outgrowth could be due to disruption of the extensive remodeling of the spore structure that accompanies germination. To evaluate this possibility, the characteristic release of DPA from the spore structure, which occurs shortly after germination initiation (25, 54), was monitored. Both the magnitude and the rate of DPA release were similar in the presence or the absence of nisin (Fig. 7A).

FIG. 7.

Effects of nisin on spore remodeling during germination. (A) At the indicated times, cultures were evaluated for the release of DPA, as described under Materials and Methods. The means of the data from a single experiment conducted in triplicate are presented. The data are representative of those from three independent experiments. Error bars indicate standard deviations. (B) After 90 min, the indicated samples were removed, fixed, and imaged by TEM, as described under Materials and Methods. RLU, relative light units.

To evaluate if nisin inhibits downstream remodeling of the spore structure, which includes hydrolysis of the cortex and release from the spore coat and exosporium. In the absence of nisin, TEM analysis revealed that the core, cortex, spore coat, and exosporium were readily evident in dormant spores (e.g., spores incubated in 0.1 M MOPS, pH 6.8) but that only the core remained in germinated spores (Fig. 7B). Nisin did not inhibit the loss of the cortex, spore coat, or exosporium in spores that had been incubated in BHI medium (Fig. 7B). Taken together, these results indicate that the nisin-mediated action against spores likely does not involve the inhibition of the extensive remodeling of the spore structure that accompanies germination.

DISCUSSION

The inhibitory action of nisin against spores of Bacillus and Clostridium pathogens has been recognized for well over 50 years (7, 36). However, the mode of action responsible for preventing spore outgrowth had not previously been characterized in detail. By using B. anthracis Sterne 7702 as a model, the data presented here demonstrate that spores lose their heat resistance and become hydrated in the presence of nisin, thereby ruling out a possible mechanism of inhibition in which nisin blocks germination initiation. Rather, germination initiation is requisite for the action of nisin. These observations are consistent with the findings of a previous report on subtilin, a close analog of nisin that also inhibits spore outgrowth without disrupting spore hydration (31). The current study revealed for the first time that nisin rapidly and irreversibly inhibits growth by preventing the establishment of oxidative metabolism and the membrane potential in germinating spores, possibly revealing an underlying explanation for the absence of B. anthracis proliferation. On the other hand, nisin had no detectable effects on the typical changes associated with the dissolution of the outer spore structures (e.g., the spore coats, cortex, and exosporium). Thus, the action of nisin reveals insights into germination by uncoupling two critical sequences of events necessary for the outgrowth of spores: the establishment of metabolism and the shedding of the external spore structures.

The capacity of nisin to prevent germinating B. anthracis spores from establishing a full membrane potential or oxidative metabolism was likely linked to the disruption of membrane integrity. Although nisin at 1 μM induced only a 6-fold increase in the amount of PI uptake above background, whereas at 100 μM nisin induced a 56-fold increase (Fig. 6), spore outgrowth and metabolic activity were still inhibited and spores were unable to establish a full membrane potential through 10 h. In contrast, ciprofloxacin, a DNA gyrase inhibitor which is recommended for use for the treatment of B. anthracis infections (8), did not prevent the establishment of a membrane potential in germinating spores and had an almost negligible effect on membrane integrity, consistent with the notion that these two antibiotics (i.e., ciprofloxacin and nisin) inhibit the cellular proliferation of B. anthracis in fundamentally different ways.

Two distinct mechanisms, membrane pore formation and the prevention of cell wall biosynthesis, contribute to the bactericidal activity of nisin against vegetative gram-positive bacteria (6, 21, 46). Our studies do not directly reveal which, if either, of these two mechanisms is primarily responsible for preventing the outgrowth of B. anthracis spores. However, nisin's capacity to disrupt the integrity of the membrane of germinating spores suggests that the membrane pore-forming activity may be important for the inhibition of spore outgrowth. In black lipid systems, nisin-induced pores allow the efflux of ATP (46), and it is conceivable that in germinating spores, the efflux of ATP through nisin-induced pores could deprive B. anthracis of the energy required for macromolecular synthesis and oxidative metabolism. Moreover, the formation of nisin-induced pores (20) can counteract the proton efflux required for membrane potential establishment and ATP formation (38, 48). Because the nisin-mediated inhibition of outgrowth requires germination initiation, its target of action likely becomes accessible only subsequent to germination initiation. It cannot currently be ruled out that nisin inhibition of cell wall biogenesis, especially at lower nisin concentrations, at which the disruption of membrane integrity is more modest, may also contribute to the prevention of spore outgrowth. Moreover, considering the structural differences between spores and vegetative bacilli, one also cannot dismiss the possibility that nisin may act upon germinating spores by a mechanism fundamentally different from that which it uses against bacilli. One study with Bacillus cereus implicated accessible thiol groups within B. cereus spores as potential targets for nisin, with the result being outgrowth inhibition (41), although a specific molecular target was not identified in that report. Prior structure-activity studies suggested that the dehydroalanine in position 5 of nisin is important for the inhibition of Bacillus spore outgrowth (9, 41), but this dehydrated residue is not essential for bioactivity in vegetative cells. In contrast, a more recent study reported that this dehydroalanine was not essential for nisin's inhibitory activity against Bacillus subtilis spores (44). In L. lactis, truncated nisin A mutants lacking rings D and E were unable to permeate the membranes or cause a disruption of the membrane potential, but these mutants retained the capacity to inhibit the outgrowth of B. subtilis spores (44). These results point to an activity other than pore formation, possibly inhibition of cell wall biogenesis, for the inhibition of spore outgrowth by nisin. Thus, structure-activity relationships for the identification of residues important for the various consequences of nisin against spores remain an important focus of future work.

Nisin is an FDA-approved natural product that has been used for 40 years for food preservation, due in part to the selective toxicity of this lantibiotic toward gram-positive bacteria (13, 15, 52). In this study, B. anthracis was used as a model, but previous work indicated that nisin is also inhibitory against spores from other Bacillus species (3, 7, 30, 32, 35, 43), as well as from Clostridium species (36), suggesting that the new information on the inhibitory activity of nisin obtained in this study will be applicable to determination of the mechanism of action of nisin against spores from these other organisms as well. Here, nisin was demonstrated to act upon and kill germinated spores of B. anthracis prior to development into elongated and dividing bacilli and before LT was generated. Notably, this mode of activity is in contrast to the modes of activity of several other widely used classes of antibiotics, including ciprofloxacin, whose mechanisms of action require ongoing cell activity and/or proliferation (2, 16, 18, 22) and which are thus not as likely to be effective against germinating spores. Collectively, these properties potentially make nisin an attractive chemotherapeutic agent for prophylaxis or postexposure treatment of spore-forming Bacillus or Clostridium pathogens.