Enterocin P Causes Potassium Ion Efflux from Enterococcus faecium T136 Cells

Abstract

Enterocin P is a bacteriocin produced by Enterococcus faecium P13. We studied the mechanism of its bactericidal action using enterocin-P-sensitive E. faecium T136 cells. The bacteriocin is incapable of dissipating the transmembrane pH gradient. On the other hand, depending on the buffer used, enterocin P dissipates the transmembrane potential. Enterocin P efficiently elicits efflux of potassium ions, but not of intracellularly accumulated anions like phosphate and glutamate. Taken together, these data demonstrate that enterocin P forms specific, potassium ion-conducting pores in the cytoplasmic membrane of target cells.

Antimicrobial peptides occur in a wide range of organisms, including bacteria, plants, and animals (10, 21). Many lactic acid bacteria produce a special kind of antimicrobial peptides or protein, the so-called bacteriocins (9, 20). Bacteriocins are peptides or proteins that kill bacteria related to the producer strain. In the struggle for a niche and nutrients, bacteriocins are useful to their producers by killing competing bacteria. Bacteriocins have been classified into three groups (19, 20): (i) lantibiotics, which contain posttranslationally modified amino acids, such as lanthionine and β-methyl-lanthionine, (ii) small (<10 kDa), heat-stable peptides without posttranslationally modified amino acids, which are subdivided into four subgroups (IIa, peptides with the N-terminal consensus sequence YGNGVXC, strongly active against Listeria spp.; IIb, two-peptide systems; IIc, sec-dependent bacteriocins; IId, class II bacteriocins not included in the previous groups), and (iii) large (>30 kDa), heat-labile proteins. Enterocin P is a bacteriocin composed of 44 amino acids produced by Enterococcus faecium P13 (5) and other related strains (11). Since it is synthesized with a cleavable signal sequence, it may be secreted via the sec system (for a review, see reference 7). Among the sec-dependent bacteriocins (14, 15, 24, 25), enterocin P is unique in having both the consensus sequence YGNGVXC and a wide inhibitory spectrum. Enterocin P is bactericidal against food-borne gram-positive pathogenic bacteria, including Staphylococcus aureus, Clostridium perfringens, Clostridium botulinum, and Listeria monocytogenes (5). Here, we studied the mode of bactericidal activity exerted by enterocin P on enterocin-P-sensitive E. faecium T136 cells.

MATERIALS AND METHODS

Materials.

⁸⁶Rb⁺ (10 mCi/mg), [¹⁴C]glutamic acid (260 mCi/mmol), and ³³P_i (3,000 Ci/mmol) were obtained from Amersham, Little Chalfont, United Kingdom. Dioleoylphosphatidylcholine (DOPC) and dioleoylphosphatidylglycerol (DOPG) were obtained from Avanti Polar Lipids. Nisin was a gift from Aplin & Barrett. The fluorescent probes 3-benzenedicarboxylic acid, 4,4′-[1,4,10,13-tetraoxa - 7,16 - diazacyclooctadecane - 7,16 - diylbis(5 - methosy - 6,2 - benzofurandliyl)] (PBFI), 3,3′-dipropylthiadicarbocyanine iodide [DiSC₃ (5)], and 2′,7′-bis-(2-carboxyethyl)-5(and 6)-carboxyfluorescein (BCECF) were obtained from Molecular Probes, Eugene, Oreg.

Strains and culture conditions.

Enterocin P-sensitive (T136s) (5) and enterocin P-resistant (T136r) strains of E. faecium T136 were used in addition to the producer strain E. faecium P13. T136r cells were isolated after selection by treating 10⁸ CFU of T136s cells/ml with 10,000 bacteriocin units (BU)/ml of enterocin P for 24 h. E. faecium T136r was not at all inhibited by an enterocin P sample that showed activity of 7,500 BU/ml against E. faecium T136s. The enterocin P resistance of E. faecium T136r remained stable after culturing in the absence of enterocin P. Both enterocin P-resistant (T136r) and enterocin P-sensitive (T136s) E. faecium T136 cells produce the enterocins A and B (4). Cells were grown in MRS broth (Oxoid) at 30°C and harvested in the logarithmic growth phase.

Enterocin P purification and antimicrobial activity assays.

Enterocin P was purified as described previously (5). The bacteriocin was dissolved in 60% (vol/vol) isopropanol and 0.1% (vol/vol) trifluoroacetic acid and stored at −20°C. An equal volume of the solvent without bacteriocin was used in control experiments. Bacteriocin activity was measured using a microtiter plate assay system (12). In brief, the optical density at 660 nm of microwell cultures exposed to a range of bacteriocin concentrations was measured.

Proton motive force measurements.

The transmembrane pH gradient (ΔpH) was measured by monitoring the fluorescence of the pH-sensitive fluorescent probe BCECF (excitation wavelength, 502 nm, and slid width, 5 nm; emission wavelength, 525 nm, and slid width, 15 nm) (17). Cells were loaded with BCECF by incubating 20 μl of 50 mM KP_i cell suspension with 1 to 3 μl of 10 mM BCECF and 2 to 2.5 μl of 0.5 N HCl for 5 min. The loading was followed by four rapid washes (Eppendorf centrifuge, 2 min at 6,000 rpm). The transmembrane electrical potential (ΔΨ) was recorded by measuring the fluorescence of 0.5 μM DiSC₃ (5) (excitation wavelength, 643 nm, and slit width, 10 nm; emission wavelength, 666 nm, and slit width, 10 nm) (22).

Uptake and efflux measurements.

E. faecium cells were harvested in the logarithmic growth phase, washed, and suspended at 137.5 μg of protein/ml in the buffers (1 to 2 ml) described in the figure legends. Uptake of radiolabeled compounds was monitored after the energization of the cells with 0.5% (wt/vol) glucose. After 20 min, either enterocin P (60 BU/ml) or solvent was added. At intervals, samples (100 μl) were applied to 45-μm-pore-size cellulose nitrate filters (Millipore Corp.) and washed twice with 2 ml of 50 mM morpholineethanesulfonic acid (MES)–NaOH (pH 7.0) (P_i efflux) or 100 mM LiCl. The radioactivity retained by the filters was measured by liquid scintillation counting in a Tri-Carb 460 CD counter (Packard Instruments Corp.).

Liposomes composed of DOPG-DOPC (1:1 [wt/wt]) and prepared by reverse-phase evaporation (23) were loaded with ⁸⁶Rb⁺ by overnight incubation at room temperature. Enterocin P (80 BU/ml) or solvent was added and efflux was measured as described above, except that filters were washed twice with 2 ml of 50 mM NaP_i, pH 7.0.

K⁺ flux measurements.

Flux of K⁺ was monitored by the K⁺-specific fluorescent indicator PBFI (excitation wavelength, 336 nm, and slit width, 15.0 nm; emission wavelength, 507 nm, and slit width, 8.0 nm) (13). Liposomes composed of DOPG-DOPC (1:1 [wt/wt]) were prepared by ethanol injection (1). Extraliposomal potassium was removed by centrifuging potassium-loaded liposomes for 15 min at 280,000 × g.

Miscellaneous methods.

Protein concentration was measured by the DC Protein Assay (Bio-Rad, Hercules, Calif.). The hydrophobicity profile of enterocin P was calculated with a 19-residue window by using the hydrophobicity scale of Eisenberg (8). Experiments were performed at 30°C and repeated at least three independent times, and typical experiments are presented.

RESULTS

Enterocin P-mediated ΔΨ dissipation.

The ability of enterocin P to dissipate ΔΨ was studied by two independent methods. First, ΔΨ was induced by energizing cells with glucose, followed by conversion of ΔpH into ΔΨ by the addition of the H⁺/K⁺ exchanger nigericin. Formation of ΔΨ resulted in a decrease in DiSC₃ (5) fluorescence. Enterocin P dissipated ΔΨ of E. faecium T136s cells in 50 mM K-HEPES (pH 7.0) (Fig. 1B) at a concentration-dependent rate (data not shown). By contrast, enterocin P-resistant E. faecium T136r cells (Fig. 1A) appeared to be completely insensitive to enterocin P action. Secondly, ΔΨ was generated by addition of the potassium ionophore valinomycin to nonenergized cells suspended in buffers without potassium. Under these conditions, enterocin P did not cause any ΔΨ dissipation (data not shown).

FIG. 1.

Enterocin P dissipates ΔΨ. Shown are enterocin P-resistant cells (T136r) (A) and enterocin P-sensitive cells (T136s) (B). To E. faecium cells (14 μg of protein/ml) suspended in 50 mM K-HEPES (pH 7.0), glucose (0.5%; arrows 1) and nigericin (250 nM; arrows 2) were added, resulting in a decrease in DiSC₃ (5) fluorescence and thus the generation of ΔΨ. Subsequently, enterocin P (80 BU/ml; arrows 3) and nisin (6 μM; arrows 4) were added, resulting (eventually) in the recovery of DiSC₃ (5) fluorescence and dissipation of ΔΨ.

Enterocin P elicits K⁺ efflux from sensitive cells.

Since enterocin P is capable of ΔΨ dissipation, it must conduct ion or proton movements across the membrane. Therefore, we investigated the ability of enterocin P to elicit transmembrane ion movements. In previous experiments (C. Herranz, unpublished data), the bacteriocin did not dissipate the ΔpH in energized cells, indicating an absence of proton conductance. To exclude that ATP-consuming proton extrusion compensated for the proton influx in these experiments, unenergized cells loaded with the pH indicator BCECF were suspended in 50 mM KP_i buffer, pH 6.5. Addition of enterocin P did not affect the BCECF fluorescence. In contrast, nigericin drastically reduced the BCECF fluorescence (data not shown). These data confirm that enterocin P does not conduct proton movements.

Subsequently, the effect of enterocin P on transport of the potassium ion analog ⁸⁶Rb⁺ was evaluated. Glucose-energized cells rapidly accumulated ⁸⁶Rb⁺ (Fig. 2). Addition of enterocin P to E. faecium T136s cells resulted in rapid (time required to deplete the intracellular concentration of radioactive substrate to 50% [t₅₀], <3 min) and drastic (more than 98%) efflux of ⁸⁶Rb⁺. In contrast, enterocin P did not cause any ⁸⁶Rb⁺ efflux from either the producer (E. faecium P13) (data not shown) or the enterocin P-resistant (E. faecium T136r) strain (Fig. 2) or large unilamellar PC-PG liposomes (data not shown). In all experiments that are represented in Fig. 2, addition of enterocin P caused the complete efflux of rubidium ions within 10 min, whereas the rubidium content of the resistant and producer cells did not decrease at all. In order to verify the potassium ion efflux, experiments were performed with the potassium ion-specific fluorescent probe PBFI. Addition of glucose to a cell suspension with external PBFI resulted in a decrease in the fluorescence due to potassium uptake (Fig. 3). Subsequent addition of enterocin P caused a reversal of the PBFI fluorescence, indicating the release of potassium ions. Synthetic PC-PG liposomes, loaded with PBFI and KP_i buffer or just with KP_i, were insensitive to enterocin P treatment at the concentrations tested (data not shown).

FIG. 2.

Enterocin P causes the efflux of ⁸⁶Rb⁺ from sensitive cells. Cells (137.5 μg of protein/ml) suspended in 50 mM NaP_i were energized with 0.5% glucose, thus allowing ⁸⁶Rb⁺ uptake. At the arrow, enterocin P (60 BU/ml) was added to E. faecium T136s (○), E. faecium P13 (▿), and E. faecium T136r (▴) cells and solvent was added to E. faecium T136s (■) cells.

FIG. 3.

Enterocin P conducts potassium ion movements in sensitive cells. E. faecium T136s cells (14 μg of protein/ml) were energized (0.5% glucose; arrow 1), after which solvent (arrow 2) or enterocin P (60 BU/ml; arrow 3) and valinomycin (0.25 μM; arrows 4) were added.

Enterocin P does not induce anion efflux.

E. faecium T136 cells treated with 60 BU/ml of enterocin P showed neither an inhibition of phosphate uptake nor efflux of accumulated phosphate. Efflux was only observed after addition of nisin (data not shown), which is known to form pores in the target membrane. A higher concentration of enterocin P (130 BU/ml) slowed down the phosphate uptake without causing any efflux (data not shown). The effect of enterocin P on transport of glutamate by E. faecium T136 cells was also determined. E. faecium T136 cells rapidly accumulated the glutamate (data not shown). A decrease in the cellular radioactivity was observed even before enterocin P or solvent addition, and this is likely due to rapid metabolization of the accumulated glutamate. There was no difference in the efflux observed between the presence of enterocin P or only solvent. On the other hand, nisin caused a significant and rapid release of the glutamate (data not shown). Enterocin P also did not inhibit uptake of glutamate (data not shown). These results suggest that enterocin P does not form aspecific pores.

DISCUSSION

Here, we investigated the bactericidal action of the bacteriocin enterocin P by studying its effect on transport processes in E. faecium T136s cells. Enterocin P causes a rapid and drastic efflux of the intracellularly accumulated potassium ion analog ⁸⁶Rb⁺ from E. faecium T136s cells. Efflux of potassium ions measured by the specific fluorescent probe PBFI shows a rapid release of intracellular K⁺. This enterocin P-mediated potassium ion efflux is highly specific, as under the same sets of conditions, no dissipation of the ΔpH was observed. Moreover, the collapse of the ΔΨ occurred only under specific conditions. No enterocin P-mediated dissipation is observed for a valinomycin-induced ΔΨ in cells suspended in Na⁺ or in choline-containing buffers. This indicates that the bacteriocin conducts neither sodium or choline ion influx. Finally, enterocin P does not cause efflux of ATP (Herranz, unpublished), radiolabeled phosphate, or glutamate.

Enterocin P has no activity at all against the producer and resistant cells nor does it seem to act on synthetic liposomes. A putative immunity protein-encoding gene is present in the producer E. faecium P13 cells (5). One might speculate that a receptor-like factor, which is absent in synthetic liposomes, might be dysfunctional in the resistant cells. However, the bacteriocin has activity against species from a variety of genera (5), which reduces the likelihood of a required cell factor.

Strikingly, enterocin P resembles the two-component bacteriocin lactococcin G in its high capacity to conduct potassium ions whereas there is no conductance of protons. In contrast to enterocin P, lactococcin G conducts a range of monovalent cations, including sodium and choline ions (18). The two-component lantibiotic lacticin 3147 dissipates ΔpH only indirectly after prolonged incubation that results in a depletion of the ATP pool (16).

Various mechanisms of membrane permeabilization by antimicrobial peptides have been proposed: the wedge-like model (6), the transmembrane helical bundle of hydrophobic peptides, the thoroidel model, the in-plane diffusion model, and the carpet model (3). Enterocin P pores may be composed of transmembrane bundles of hydrophobic peptides. This is supported by the high ion specificity of enterocin P and by the presence of a highly hydrophobic transmembrane segment in the C-terminal part of enterocin P (Fig. 4). Since the hydrophobicity increases in the N-to-C-terminal direction, the C-terminal segment of enterocin P may possibly insert into the membrane, whereby the C-terminal histidine reaches the cytoplasm. In this respect, a C-terminal intracellular histidine has been shown to determine the activity of a potassium channel (2).

FIG. 4.

Hydrophobicity profile of residues 27 to 44 of enterocin P. tms, transmembrane segment.

Taken together, bundles of transmembrane enterocin P peptides may form a pore that specifically conducts potassium ions. Future reconstitution of potassium ion-conducting enterocin P activity in liposomes may reveal the molecular interactions that underlie the observed high ion specificity of enterocin P pores.