Enterocin P Selectively Dissipates the Membrane Potential of Enterococcus faecium T136

Abstract

Enterocin P is a pediocin-like, broad-spectrum bacteriocin which displays a strong inhibitory activity against Listeria monocytogenes. The bacteriocin was purified from the culture supernatant of Enterococcus faecium P13, and its molecular mechanism of action against the sensitive strain E. faecium T136 was evaluated. Although enterocin P caused significant reduction of the membrane potential (ΔΨ) and the intracellular ATP pool of the indicator organism, the pH gradient (ΔpH) component of the proton motive force (Δp) was not dissipated. By contrast, enterocin P caused carboxyfluorescein efflux from E. faecium T136-derived liposomes.

Bacteriocins produced by bacteria are a heterogeneous group of ribosomally synthesized antimicrobial proteins, which display antimicrobial activity against other bacteria (16, 18, 29). The genus Enterococcus is among the lactic acid bacteria (LAB) associated with foods that produce bacteriocins (2, 4, 6, 11, 12). Enterococcus faecium P13 and other E. faecium strains isolated from dry-fermented sausages produce enterocin P, a bacteriocin with strong inhibitory activity against Listeria monocytogenes (11, 14). Enterocin P is a pediocin-like bacteriocin (27) with an N-terminal signal peptide which may allow it to be secreted via the sec pathway (for a review, see reference 30). The enterocin P operon has been previously characterized (11), and it consists of the bacteriocin structural gene (entP) encoding a 71-amino-acid precursor with a 27-amino-acid signal peptide and has a second open reading frame (orf2) located immediately downstream of entP and potentially encoding the immunity protein.

The increasing interest of consumers in minimally processed, naturally preserved foods has prompted the proposal to use bacteriocinogenic LAB or their bacteriocins as biopreservatives to increase microbiological safety (19, 25, 31). However, rational application of the bacteriocins requires an understanding of the mechanisms underlying their specificity and activity (24, 26), as well as knowledge of the structure-function relationships, to develop new compounds with an improved efficacy. In this context, the objective of this study was to examine the mechanistic action of enterocin P on the sensitive strain E. faecium T136.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

E. faecium P13 (11) and E. faecium T136 (6) were propagated in MRS broth (Difco Laboratories, Detroit, Mich.) at 30°C and used as the enterocin P producer and indicator microorganisms for bacteriocin activity, respectively.

Purification of enterocin P.

The bacteriocin was purified from the supernatant of a 16-h E. faecium P13 culture by hydrophobic interaction and cation-exchange chromatographies, basically as described by Casaus et al. (6). Next, the sample was dialyzed against distilled water in dialysis tubing (Spectrum, Houston, Tex.; molecular weight cutoff, 1,000) and concentrated with polyethylene glycol to half of its original volume. This sample was further purified by preparative isoelectric focusing using a Rotofor cell (Bio-Rad Laboratories, Melville, N.Y.), as previously described (8, 10). Fractions displaying the highest antimicrobial activity were pooled together, subjected to ultrafiltration (Centricon-3 concentrators; Amicon; molecular weight cutoff, 3,000), and stored at 4°C until used.

Activity assays.

The antimicrobial activity of the fractions obtained throughout the purification process was calculated by a microtiter plate assay performed basically as described by Holo et al. (15). Briefly, 96-well plates containing 50 μl of twofold diluted fractions and 150 μl of the freshly diluted indicator organism (2.5 × 10⁶ CFU ml⁻¹) per well were incubated at 30°C for 16 h. Growth inhibition was measured spectrophotometrically at 620 nm with a microtiter plate reader (Dynatech MR5000; Dynatech Laboratories), and bacteriocin activity was calculated in bacteriocin units (BU). One BU was defined as the reciprocal of the highest dilution of bacteriocin causing 50% growth inhibition (50% of the turbidity of the control culture without bacteriocin).

Measurements of proton motive force.

The membrane potential (ΔΨ) of E. faecium T136 cells was qualitatively measured with the fluorescent probe 3,3′-dipropylthiadicarbocyanine iodide [(DiSC₃(5)] (Molecular Probes Inc., Eugene, Oreg.). Cells were harvested in the log phase (optical density at 660 nm [OD₆₆₀], 0.6), washed twice with ice-cold 50 mM potassium HEPES (K-HEPES) buffer, pH 7.0, resuspended in the same buffer to 1/100 of their initial volume, and stored on ice. Glucose-energized E. faecium T136 cells (final OD₆₆₀, 0.3) were added to a stirred cuvette containing 2 ml of the K-HEPES buffer and DiSC₃(5) (5 μM). Next, nigericin (1.5 nM), which dissipates the pH gradient (ΔpH), and enterocin P (200 BU/ml) or valinomicyn (1.5 nM) were added. Fluorescence measurements were performed with a F1T11 spectrofluorometer (Spex Industries, Metuchen, N.J.) with a band-pass width of 10 nm and wavelengths of 643 and 666 nm for excitation and emission, respectively.

The transmembrane ΔpH was measured by loading E. faecium T136 cells (OD₆₆₀, 0.6) with the fluorescent probe 2′,7′-bis-(2-carboxyethyl)-5[and 6]-carboxyfluorescein acetoxymethyl ester (BCECF AM) (Molecular Probes Inc.) by using an acid shock, as described by Molenaar et al. (21). Glucose-energized, BCECF-loaded cells (final OD₆₆₀, 0.15) were added to a stirred cuvette containing 2 ml of 50 mM KP_i buffer, pH 6.0. Next, valinomycin (1.5 nM), which dissipates the ΔΨ, and enterocin P (200 BU/ml) or nigericin (1.5 nM) were added. Fluorescence was measured with band-pass widths of 5.0 and 15.0 nm and wavelengths of 643 and 666 nm for excitation and emission, respectively.

Measurement of ATP.

E. faecium T136 cells grown to an OD₆₆₀ of 0.6 were collected by centrifugation, washed once with 50 mM 2-(N-morpholino)ethanesulfonate (MES) buffer, pH 6.5, and kept on ice until use. In order to energize the cells prior to ATP measurements, they were resuspended to half of the original volume in 50 mM MES buffer with 0.2% glucose and incubated for 20 min. After treating the cells with 5, 20, 50, and 100 BU of enterocin P/ml, total and external cellular ATP levels were determined using the bioluminescence method described by Chen and Montville (7) and an ATP bioluminescence assay kit (Sigma Chemical Co., St. Louis, Mo.). Intracellular ATP was calculated by subtracting external from total ATP. The assays were calibrated by using a standard curve obtained by measuring the bioluminescence of ATP solutions of known concentrations, and ATP levels were expressed as nmol mg⁻¹ of cells (dry weight). Values are the mean of two independent bioluminiscence measurements.

Lipid extraction, preparation of CF-loaded liposomes, and CF leakage assay.

E. faecium T136 cells were grown to mid-log phase (OD₆₆₀, 0.7 to 0.8), harvested, and washed with 0.1% peptone water. Total E. faecium T136 lipids were extracted following the procedure of Bligh and Dyer (as described by New in reference 28) with the modifications introduced by Winkowski et al. (35). Extracted lipids were kept at −20°C in glass vials and were used within 2 weeks. Next, large unilamellar vesicles composed of lipids from E. faecium T136 and loaded with 6-carboxyfluorescein (CF) (Sigma Chemical Co.) were prepared as described by Chen et al. (8). CF-loaded liposomes were stored on ice and used within 3 h.

The effect of enterocin P (90 and 450 BU/ml) on CF-loaded E. faecium T136-derived liposomes was determined by monitoring the fluorescence of the liposomal suspension upon bacteriocin addition. Fluorescence measurements were performed with a band-pass width of 0.8 nm and wavelengths of 516 and 490 nm for emission and excitation, respectively. Results were expressed as the percentage of CF release, calculated from the following equation: % efflux = (F_t−F₀)/(F_∞−F₀), where F_t was the fluorescence at time t, F₀ was the control fluorescence at time t, and F_∞ was the fluorescence after the addition of Triton X-100 (10% [vol/vol] in distilled H₂O). The fluorescence values were corrected by subtracting the fluorescence of the control samples treated with 50 mM MES buffer instead of bacteriocins.

RESULTS

Effect of enterocin P on ΔΨ and ΔpH.

The effect of enterocin P on the ΔΨ was qualitatively determined as quenching of the fluorescent probe DiSC₃(5) from the cells treated with the bacteriocin (Fig. 1). After glucose addition, an increase in the fluorescence occurred as a result of the net proton extrusion by the membrane-bound F₀F₁-ATPase. Upon addition of enterocin P to energized, nigericin-treated cells, an increase in fluorescence intensity was observed, which was an indication of a decrease in internal membrane potential. After a short initial period, the fluorescence increase induced by enterocin P was greater than that provoked by valinomycin.

FIG. 1.

Effect of enterocin P (200 BU/ml) (a) and valinomycin (1.5 nM) (b) on the ΔΨ of E. faecium T136 cells. Fluorescence levels before the addition of enterocin P or valinomycin were arbitrarily designated zero, and the increase in fluorescence upon the addition of bacteriocin or ionophore was expressed in arbitrary units (a.u.).

Changes in intracellular pH caused by the addition of enterocin P to the cellular suspension were qualitatively recorded by measuring the fluorescence of the probe BCECF AM (Fig. 2). After the cells were energized, the addition of valinomycin dissipated the ΔΨ and resulted in a fluorescence increase that was maintained until the addition of nigericin, which rapidly reversed the pH to a level equal to that of the medium pH. The addition of 200 BU of enterocin P ml⁻¹ did not dissipate ΔpH. Furthermore, an increase in the pH gradient was observed upon bacteriocin treatment, which was more pronounced in cells that had not been previously treated with valinomycin. Higher enterocin P concentrations (i.e., 350 BU ml⁻¹ [results not shown]) produced similar behavior, i.e., a transient increase preceding the fluorescence decrease.

FIG. 2.

Effect of enterocin P (200 BU/ml) on the ΔpH of valinomycin-treated (a) or untreated (b) E. faecium T136 cells and of nigericin (1.5 nM) (c). Fluorescence levels before the addition of enterocin P or nigericin were arbitrarily designated zero, and the variation in fluorescence upon the addition of bacteriocin or ionophore was expressed in arbitrary units (a.u.).

Enterocin P depleted intracellular ATP levels of E. faecium T136 cells.

Upon energization, ATP levels of E. faecium T136 cells increased to a mean value of 7.2 nmol mg⁻¹ of cells (dry weight). This value was similar to that reported by Chen and Montville (7) for energized L. monocytogenes cells. Enterocin P depleted intracellular ATP in a time- and concentration-dependent manner (Fig. 3). After 20 min of treatment, ATP levels were 39, 14, and 0.2% of the original ATP concentrations for cells that had been treated with 5, 20, and 50 BU ml⁻¹, respectively. Bacteriocin concentrations higher than 50 BU ml⁻¹ (i.e., 100 BU ml⁻¹; data not shown) did not induce any further intracellular ATP depletion. No significant appearance of ATP was found in the external medium at any of the enterocin P concentrations tested.

FIG. 3.

Intracellular ATP levels of E. faecium T136 cells treated with 5 (●), 10 (▾), and 50 (■) BU of enterocin P/ml and untreated (⧫). No extracellular ATP was detected (data not shown).

Enterocin P-induced CF efflux from E. faecium T136-derived liposomes.

Exposure of E. faecium T136-derived liposomes to enterocin P caused a gradual release of CF in a concentration- and time-dependent fashion (Fig. 4). CF efflux plateaued earlier at the lower enterocin P concentration (90 BU ml⁻¹). After a 600-s treatment, the percentages of CF efflux were 19 and 33% for 90 and 450 BU ml⁻¹ of enterocin P, respectively.

FIG. 4.

Effect of 90 (a) and 450 (b) BU of enterocin P/ml on E. faecium T136-derived liposomes. The CF efflux is expressed as the percentage of the release induced by Triton X-100 (0.5%, wt/vol).

DISCUSSION

It is generally accepted that LAB bacteriocins act by altering the permeability barrier of the cell membrane (13, 22, 33, 34, 35) and that one of the common mechanisms of inhibition of their target cells is the dissipation of the proton motive force (PMF) (1, 5, 25). Since enterocin P is an amphipathic, cationic peptide (11), it may interact with the negatively charged bacterial membranes of the sensitive cells and alter their properties. In order to check this hypothesis, the components of the PMF were measured in E. faecium T136 cells treated with enterocin P. Enterocin P efficiently dissipated the ΔΨ of nigericin-treated cells; the fact that the bacteriocin-induced ΔΨ dissipation was greater than the valinomycin-induced one may indicate that mechanisms different from the simple outward diffusion of the dye in response to the decrease in ΔΨ could be involved, for example, several degrees of membrane disruption.

When the same concentration of enterocin P was used, ΔpH dissipation of valinomycin-treated or untreated E. faecium T136 cells was not observed. Moreover, a slight increase in ΔpH was recorded, which was greater in cells not treated with valinomycin. This contrasts with the observation that, generally, the ΔpH component of the PMF is dissipated earlier than the ΔΨ one (5, 17, 32). However, like enterocin P, lactococcin G, a monovalent cation-conducting bacteriocin (22, 23), and lacticin 3147 (20) exert a selective dissipation of ΔΨ and may cause an increase in ΔpH. Moll et al. (23) explained this effect by the enhanced H⁺ extrusion by the F₀F₁-ATPase as a consequence of ΔΨ dissipation. In the case of lacticin 3147, a slow decrease in ΔpH was observed after its increase. This may be a secondary effect of the diminished intracellular ATP pool, which was depleted in an attempt to maintain the PMF (20).

The energetic state of the cells treated with enterocin P was evaluated by measuring their ATP levels. The addition of enterocin P to energized E. faecium T136 cells induced intracellular ATP depletion without ATP efflux. This effect has been reported for several class II bacteriocins, such as lactacin F (1), pediocin PA-1 (7), lactococcin G (22), and mundticin (4). The ATP negative charge and size (M_w, 507) may play a role in determining its inability to pass through the pore. The fact that internal ATP may be depleted in the absence of ATP efflux suggests that it is hydrolyzed inside the sensitive cells. Two mechanisms have been proposed to explain the hydrolysis: (i) a shift in the equilibrium of the ATP hydrolysis reaction as a consequence of P_i loss through the membrane with impaired permeability (1) or (ii) accelerated hydrolysis due to the cell's attempt to regenerate the electrochemical gradient by H⁺ extrusion driven by the F₀F₁-ATPase energy-consuming pump (7, 23).

Finally, the effect of enterocin P in E. faecium T136-derived liposomes was determined. The bacteriocin induced CF efflux in liposomes in a time- and concentration-dependent fashion and in the absence of a PMF. CF release showed saturation kinetics and plateaued far from 100% efflux. This suggests that the permeabilization effect of enterocin P is transient in contrast with the “all or none” disruptive action of detergent peptides. The bacteriocin acts on liposomes in an energy-independent fashion.

The fact that enterocin P was able to act in a liposomal, protein-free system indicated that a receptor of a proteinaceous nature is not absolutely essential for bacteriocin action. The requirement of a receptor-like factor to exert antimicrobial activity has been suggested for the lactococcins A (33), B (34), and G (22) and some other class II bacteriocins. Although the broad antimicrobial spectrum of enterocin P (11) supports the evidence that a protein receptor is not a requirement for the activity, the activity enhancement in vivo by some proteinaceous components of the membrane or cell surface, or even the whole membrane constitution, should not be discarded (3, 9, 22).

Further studies will establish the ion specificity and the molecular mechanism of pore formation for enterocin P, including energy requirement and lipid dependence.