Class IIa bacteriocins from lactic acid bacteria: Antibacterial activity and food preservation

OF BIOSCIENCE AND BIOENGINEERING Vol. 87, No. 6, 705-716. 1999 JOURNAL

REVIEW Class IIa Bacteriocins from Lactic Acid Bacteria: Antibacterial Activity and Food Preservation SAID ENNAHAR,

KENJI SONOMOTO,

AND

AYAAKI

ISHIZAKI*

Laboratory of Microbial Science and Technology, Division of Bioscience and Biotechnology, Graduate School of Bioresource and Bioenvironmental Sciences,Kyushu University, 6-10-l Hakozaki, Higashi-ku, Fukuoka 812-8581, Japan Received 14 April 1999/Accepted 22 April 1999

In the last decade, a variety of rlbosomally synthesized antimicrobial peptides, or bacterlocins, produced by lactic acid bacteria have been identified and characterized. As a result of these studies, insight has been gained into various fundamental aspects of biology and biochemistry such as bacteriocin processing and secretion, mechanisms of cell immunity, and structure-function relationships. In parallel, there has been a growing awareness that bacterlocins may be developed into useful antimicrobial food additives. Class IIa bacterlocins can be considered as the major subgroup of bacteriocins from lactic acid bacteria, not only becauseof their large number, but also because of their significant biological activities and potential applications. The present review provides an overview of the knowledge available for classIIa bacteriocins and discussescommon features and recent findings concerning these substances. The activity and potential food applications of classIIa bacteriotins are a major focus of this review. [Key words: bacteriocin,

class IIa bacteriocins,

lactic acid bacteria, Listeria, food preservation]

In recent years, numerous food poisoning outbreaks, implicating various pathogens and food products, and the increasing concern over the preservation of minimally processed foods have spurred growing awareness of the importance of food safety (1). This has prompted new approaches to inhibit foodborne pathogens; in particular there has been a renewed interest in lactic acid bacteria (LAB), whose antimicrobial activity has been used for centuries to preserve food, for production of bacteriocins. As a result of intense investigations on LAB bacteriocins, considerable progress has been made in both basic and applied research disciplines towards a better understanding of these substances and a large number of chemically diverse bacteriocins have been identified. Four classes of bacteriocins have then been defined based on observed common characteristics, mainly structural (2). New bacteriocins are still being discovered and regularly reviewed and documented in books and reviews, with class I (Iantibiotics) and class II (small heat-stable non-lanthionine-containing peptides) bacteriotins being the most abundant and thoroughly studied (3-5). As a consequence of recurring and serious listeriosis outbreaks (6, 7), Listeria monocytogenes has come under the focus of bacteriocin investigators during the past decade. The search for bacteriocin-producing LAB has then been directed towards substances whose targets are the Listeria spp., and consequently this has led to the description of a large number of antilisterial bacterio-

tins. These belong to either class I or II, but recent reports clearly showed that they are predominantly class IIa bacteriocins, a subgroup of bacteriocins classified on the basis of their strong amino acid sequence similarity, particularly their distinctive N-terminal part, and their strong antilisterial activity (2, 8). In fact, all class IIa bacteriocins identified so far are particularly strong inhibitors of the pathogen Ls. monocytogenes (8-22). Because of this antilisterial effectiveness, class IIa bacteriotins have significant potential as biopreservatives in a large number of foods. Moreover, several class IIa bacteriocins appear to have potential applications against other spoilage and foodborne pathogenic microorganisms, since they display spectra of activities that may be broad enough to encompass undesirable bacteria, such as spoilage LAB, Brochothrix spp., Clostridium spp., Bacillus spp., and Staphylococcus spp. (9, 10, 23). They are now the most promising bacteriocin candidates for various industrial applications, not only due to their biological activity, which is generally higher than that of other bacteriocins (9), but also because of their physicochemical properties. Class IIa bacteriocins, also called pediocin-like bacteriocins after the first discovered member (pediocin AcH/PA-l), can be considered the largest (with 14 members) and the most extensively studied subgroup of LAB bacteriocins. The present review provides an overview of knowledge available for these bacteriocins, with a particular focus on their biological activities and potential as preservatives for food.

* Corresponding author. Abbreviations: C., Carnobacterium; GMPs, good manufacturing practices; GRAS, generally recognized as safe; HACCP, hazard analysis critical control point; L., Lactococcus;LAB, lactic acid bacteria; Lb., Lactobacillus;Ls., Listeria; P., Pediococcus;PI, isoelectric point; PMF, proton motive force; A$, transmembrane potential; X,,, variable amino acid residue; ApH, pH gradient.

GENERAL

NATURE OF CLASS IIa BACTERIOCINS

Composition and structure Class IIa bacteriocins that are identified so far contain between 37 residues (leucocin A and mesentericin Y105 [also called Y 10537, 705

706

ENNAHAR

ET AL.

J. BIOSCI. BIOENG.,

although the term Y105 is used in this review]) and 48 residues (carnobacteriocin B2) and they are considerably homologous with each other (Fig. 1). They were first described based on the presence in their N-terminal halves of the YGNGVX,,C motif, which has been suggested to be part of a recognition sequence for a speculated membrane-bound protein receptor (24, 25). Except for leucocin A and mesentericin Y105, which vary only in the amino acids at positions 22 and 26, class IIa bacteriocins share up to 80.5% sequence identity, e.g., as in sakacin A and enterocin P. Additionally, several independently investigated class IIa bacteriocins have been shown to be identical to others. These include carnobacteriocin BMl and piscicocin Vlb; piscicolin 126 and piscicocin Vla; sakacin A and curvacin A; pediocin PA-l, pediocin AcH and pediocin SJ-1 (Fig. 1). Another common feature of class IIa bacteriocins is their net positive charge, with pIs varying from 8.3 to 10, thereby showing close net charges at various pH values. However, this seems to be a common feature for all bacteriocins (3). Class IIa bacteriocins present a number of other close characteristics relating to the presence of particular groups of residues: a high content of amino acid residues with ionizable side chain groups, specially basic amino acids, and also nonpolar residues and small amino acids such as glycine, which is thought to confer to these bacteriocins a high degree of conformational freedom (15). The highly conserved N-terminal hydrophilic domain clearly contrasts with the moderately conserved hydrophobic and/or amphiphilic C-terminal domain, which relates to the bacteriocin activity on target membranes (26). With regard to their predicted secondary structures, class IIa bacteriocins exist primarily as unstructured conformations, generally random coils, in aqueous solutions, whereas in non aqueous solutions they adopt a partly helical structure with varying degrees of hydrophobicity and other defined secondary structures (2, 3, 25, 27, 28). The N-terminus of class IIa bacteriocins is predicted to contain p-sheets maintained in a /?-hairpin conformation that is stabilized by the N-terminal disulfide bridge. This conformation is believed to give class IIa bacteriocins an amphiphilic characteristic in the Nterminal region (29, 30). As for the C-terminal half, it has been predicted to adopt an amphiphilic a-helix, spanning similar regions in different molecules (2, 9, 12, 24, 25, Bacteriocin Leucocin A MesentericinY105 Mundticin PisciwIin 126 Bavaricin A salracii P Pediocin PA-l Bavaricin MN Diver& V41 EntemcinA TTH Enter&n P CamobacteriricinBMl SakaciA Camoba&riocin B2

-

30, 31), which are believed to be the transmembrane segments during pore formation in sensitive cell’s membrane (2, 3, 19, 26). In addition, class IIa bacteriocins are cystibiotics, i.e. they have at least two cysteines with disulfide bridges. In the alignment of class IIa bacteriocins (Fig. l), it appears that the two cysteine residues in the N-terminal domain are present at conserved positions, and consequently the disulfide bridge which forms a six-membered ring over these two residues is well conserved in all class IIa bacteriocins. Moreover, pediocin PA-l/AcH, enterotin A and divercin V41 are unique in the sense that they possess an extra disulfide bond involving a second pair of cysteine residues (Fig. 1). The presence of disulfide bonds seems to be crucial for the activity of class IIa bacteriocins, particularly those with two disulfide bridges (23). Occurrence among LAB bacteriocins Class IIa bacteriocins are produced by various LAB, including Lactobacillus spp . , En terococcus spp . , Pediococcus spp . , Leuconostoc spp., and Carnobacterium spp. Interestingly, unlike other LAB bacteriocins, all class IIa bacteriotins are produced by food-associated strains, isolated from a variety of food products of industrial and natural origins, including meat products, dairy products and vegetables (Table 1). However, it is clearly apparent that most of the class-IIa-bacteriocin-producing LAB have been isolated from meat products. In addition, to date, no class IIa bacteriocin has been shown to be naturally produced by a bacterium other than the LAB, as opposed to bacteriocins of other classes, like lantibiotics (5). These are two significant reasons for further investigating the potential use of class IIa bacteriocins as food biopreservatives. Some class-IIa-bacteriocin producers isolated recently have been shown to produce more than one bacteriocin, a phenomenon that may be quite common among bacteriocin-producing LAB strains. The bacteriocins produced may belong to the pediocin family, e.g. carnobacteriocin BMl and B2 produced by C. piscicola LV 17B, and piscicocin Vla and Vlb produced by C. piscicola Vl (Table l), with only slight differences or no difference at all between their inhibitory spectra (8, 12). One of the bacteriocins produced may belong to another bacteriocin group, like enterocin B, a non class IIa bacteriotin that is produced along with the class IIa enterocin A

Audio acid sequence IK Y Y GNG V]H~TK KYY GNG VHCTKS KYY GNG VSCNKK KYY GNG VSCNKN GNG VHcGKH GNG VHCGKH GNG VTCGKH GNG VYCNSK GNG VYCNSK GNG VYCTKN GNG VYCNNS 'GNG VYCNKE 'GNG VYCNNK VNb X-!Ul ASCSKT ii

Ref. (13) (25) (19) (10) w

s

(21) I:;; (19 (11) (9) (8) GRRP -___-

(20)

_- (8)

FIG. 1. Sequence alignment of class IIa bacteriocins on the basis of the N-terminal YGNGV consensus motif. White boxes enclose residues conserved in at least 12 of the 14 sequences shown. Shaded boxes enclose C-terminal residues conserved in at least 2 sequences. Amino acid residues shown in: boldface are positively charged; small capitals are uncertain residues; and Xs are unknown residues. Sequences of the following bacteriocins are respectively identical: piscicolin 126 and piscicocin Vla (12); carnobacteriocin BMl and piscicocin Vlb (12); sakacin A and curvacin A (40); leucocin A and leucocin B-Talla (97); pediocin PA-l, pediocin AcH (95) and pediocin SJ-1 (52).

VOL.

707

CLASS IIa BACTERIOCINS OF LACTIC ACID BACTERIA

87, 1999

TABLE 1. Lactic acid bacteria producing class IIa bacteriocins and their sources Bacteriocin Leucocin A/B-Tal

Producer

la

Mesentericin Y 105 Divercin V41 Carnobacteriocin B2 Carnobacteriocin BMl Piscicocin Vl b Piscicocin Vla Piscicolin 126 Mundticin Enterocin A Enterocin P Pediocin PA-l/AcH/SJ-1

Strain UAL187 Leuconostocgelidum carnosum Tal la mesenteroides Y105 Carnobacteriumdivergens v41 LV17B piscicola LV17B Vl Vl JG126 Enterococcusmundtii AT06 faecium CTC4921T136 P13

Pediococcusparvulus acidilactici

AT0341AT077 H PAC 1.0 SJ-1

Curvacin A Sakacin A Sakacin P Bavaricin A Bavaricin MN

Lactobacihs plantarum curvatus sake

WHE92 LTHl174 Lb706 Lb674 MI401 MN

by Enterococcus faecium CTC492, the two bacteriocins displaying different inhibitory spectra (23, 32). Another example of multiple-bacteriocin production is represented by Lb. sake LTH673, a sakacin P producer, which produces at least one additional bacteriocin not yet characterized (33). Conversely, LAB strains of different origins have been shown to produce the same class IIa bacteriocin (Table l), e.g. the carnobacteriocin BMl-piscicocin Vlb-producers C. piscicola LVl7B and Vl, isolated from fresh pork and fish, respectively (8, 12); the piscicocin Vlapiscicolin 126-producers C. piscicola Vl and JG126, isolated from fish and ham, respectively (10, 12); and the pediocin PA-1-AcH-producers P. acidilactici PACl .O, H and SJ-I, isolated from fermented sausage in different countries (34-36). A pediococcal plasmid carrying the pediocin-AcH genetic information has in fact been reported to undergo conjugal transfer in matings within P. acidilactici (37). But more importantly, pediocin PAl/AcH is also naturally produced by P. parvulus strains isolated from minimally processed vegetables (38) and by the totally unrelated Lb. plantarum WHE92 isolated from cheese (39). Likewise, sakacin A and curvacin A are identical bacteriocins despite being produced by different Lactobacillus species (Table 1) (17, 20, 21, 40, 41). Additionally, leucocin A and mesentericin YlOS, which vary only in the amino acids at positions 22 and 26, are produced by different Leuconostoc species. Furthermore, bacteriocins like sakacin A, carnobacteriocin BMl and enterocin P, although produced by bacteria of three different genera, show very high degrees of homology (up to 80.5% identity), especially toward the N-terminal domain (Fig. 1, Table 1). In contrast, bacteriocins produced by different strains of the same species may show much lower homology. For instance, whereas sakatin A is identical to curvacin A (from Lb. curvatus), it has low homology with sakacin P, which is produced by a different strain of the same species, Lb. sake (20, 40, 41). These observations suggest that production of class IIa bacteriocin, while confined to LAB, seems to be quite mobile (genetic transfer would be quite common among LAB strains) and no class IIa primary structure

Source Vacuum-packaged processed meat in Canada Vacuum-nackaged processed meat in South Africa Goat’s milk in France Fish viscera Fresh pork packaged in modified atomosphere Fresh pork packaged in modified atomosphere Fish Fish Spoiled ham Fresh chicory endive Spanish dry fermented sausage Spanish dry fermented sausage Fresh chicory endive Fermented sausage Naturally-fermented sausage in Israel Soft cheese in France Fermented sausage Raw meat Meat Sour doughs Meat

Ref. (96) (97)

(22) (18) (102) (102) (98) (981 (10) (19) (11) (91 (381 (34) (351 (36) (391 (4) (101) (991

(16) (100)

seems to be distinctive of a particular LAB genus or species. ANTIBACTERIAL ACTIVITY BACTERIOCINS

OF CLASS IIa

Bactericidal effects Like other LAB bacteriocins, class IIa bacteriocins are bactericidal peptides which act primarily by permeabilizing the membranes of susceptible microorganisms, causing leakage of cellular contents (3, 26, 29, 42-46). Membrane permeabilization follows bacteriocin-membrane interaction, which is thought to involve the formation of water-filled membrane channels through a multistep process of binding, insertion and aggregation of monomers in the membrane leading to formation of poration complexes (2). Although few class IIa bacteriocins have been investigated, it is very likely that all members of the class IIa family act similarly on sensitive cells. Pore formation by class IIa bacteriocins in indicator cells, generally Listeria strains, is believed to originate a membrane ionic imbalance, mainly through efflux of potassium, and leakage of inorganic phosphate. This has been convincingly demonstrated for pediocin PA-l (42, 44, 47), mesentericin Y105 (46), and bavaricin MN (15). A consequence of such disruptions is the dissipation of proton motive force (PMF), which involves the partial or total dissipation of either or both the transmembrane potential (A#) and the pH gradient (ApH) (30). Unlike lantibiotics, which totally dissipate both A$ and ApH (30), class IIa bacteriocins readily provoke a total dissipation of ApH, but only a partial dissipation of A$ (15, 42, 43, 46). Only the newly discovered mundticin has been shown to cause a complete dissipation of A$ (1% The lethal activity of class IIa bacteriocins, is thus mainly ascribed to the dissipation of the PMF (3, 26, 48), whose role is fundamental for energy-consuming processes such as ATP synthesis and phosphate-bonddriven transport systems (49). Particularly, the intracellular ATP is depleted by rates of up to 98.9% (19, 47) and the uptake of amino acids, which is mediated by active

708

ENNAHAR

ET AL.

transport, is blocked (44, 46). Moreover, leakage of preaccumulated amino acids, among other UV-absorbing materials, has been reported for pediocin PA-l (42, 44) and mesentericin Y105 (46). The efflux of amino acids caused by these two bacteriocins may occur by diffusion through the bacteriocin pores, probably combined with reflux via PMF transport systems (30). However, the very rapid efflux of amino acids due to the action of mesentericin Y105 suggests that this resulted from simple leakage (46). Unlike lantibiotics (50, 51), no leakage of ATP seems to be caused by class IIa bacteriocins (19, 47), which is thought to be due to smaller pore sizes formed by the latter than by the former. The observed depletion of intracellular ATP may therefore result from an accelerated consumption of ATP in order to maintain or restore PMF and/or the inability of the cell to produce ATP due to phosphate efflux. By comparing the rates of pediocin-PA-l-induced ATP depletion and inorganic phosphate efflux, Chen and Montville (47) have suggested that the observed depletion of ATP is most likely due to attempts of the cell to regenerate the decreased PMF, rather than a shift in the ATP hydrolysis equilibrium due to the loss of inorganic phosphate. Again, contrasting results have been reported for the lantibiotic nisin Z (50). Models for bacteriocin-membrane interactions LAB bacteriocins, including class IIa bacteriocins, are generally active against protoplasts derived from cells of resistant gram-positive strains, which indicates that the bacteriocin resistance of some gram-positive strains is due to the barrier properties of their cell walls (52). As for how these bacteriocins cross the cell wall and reach the membrane surface of sensitive cells, readers may refer to the review of Jack et al. (3), where this feature has been extensively dealt with. Formation by class IIa bacteriocins of poration complexes in target cell membranes has been clearly demonstrated. In addition, the presence in these bacteriocins of amphiphilic segments, which are putative transmembrane helices, their water solubility and membrane-binding ability, suggest that they may form poration complexes following a barrelstave model. Understanding the mechanisms of bacteriocin-membrane interactions constitutes a major focus of current bacteriocin research, and though no conclusive evidence has yet been found, considerable progress is being made towards elucidating how each step of the multistep process leading to pore formation proceeds. The initial binding step of class IIa bacteriocins to the membrane surface is believed to be electrostatic in nature and to be mediated by a putative membrane-bound protein receptor (26, 44, 48). The involvement of such a protein receptor in membrane interactions of class Ha bacteriocins has been generally believed to be a requirement (3, 44, 48), which clearly contrasts with the receptor-independent mechanism through which lantibiotics achieve these interactions (5, 53). Nonetheless, Maftah et al. (46), who have shown that mesentericin YlO5 was able to induce pore formation in the energy-transducing membranes of the mitochondria, suggested that the receptors involved in the primary binding of mesenteritin Y105, which would not be specific for bacteriocins since they may also exist in mitochondria, could be proteins or phospholipids. More interestingly, recent studies investigating pore formation by class IIa bacteriocins in lipid vesicle systems devoid of membrane proteins

J BIOSCI. BIOENG..

indicate that protein receptors may not be absolutely required (15, 29). For instance, Chen et al. (29) have shown that pediocin PA-l was capable of permeabilizing both lipid vesicles derived from Ls. monocytogenes cells and pure phospholipid vesicles, thereby demonstrating that a protein receptor is not an absolute requirement for pore formation by this bacteriocin. This is in direct contrast to earlier findings suggesting a protein-receptormediated activity of pediocin PA-l on pediococcal cells (42, 44). Similarly, using Listeria-derived lipid vesicles, protein receptors have been shown to be unnecessary for pore formation by another class Ha bacteriocin, bavaritin MN (15). However, the possible involvement of protein receptors has not been completely ruled out, since some researchers believe that class IIa bacteriocins might function in a protein-receptor-enhanced fashion (29, 30). Therefore, it has been suggested that functional binding of the positively charged and polar residues of pediocin PA-l, bavaricin MN, and mesentericin Y105 occurs primarily in conjunction with anionic phospholipid head groups in the membrane (15, 29, 46, 54, 55). Considering the cationic nature of class IIa bacteriocins and the high similarity in structure of their hydrophilic Nterminal half (Fig. l), which is thought to mediate the initial binding, it is most likely that all class Ha bacteriotins rely at least in part on the same type of functional binding. As a subsequent step, hydrophobic interactions would occur between a hydrophobic and/or amphiphilic domain within the C-terminal half of the bacteriocin and the lipid acyl chains, and have been shown to be crucial for the pore formation process (15, 25, 54, 56). In fact, the C-terminal half of class IIa bacteriocins, which is more hydrophobic and less polar than the N-terminal half, contains a domain which appears to be involved in hydrophobic interactions with the membrane (24, 56). It has been suggested that this domain may be the cellspecific region for class IIa bacteriocins, in contrast to the N-terminal domain which interacts electrostatically with the membrane surface probably in an unspecific manner (24, 56). In such a case, hydrophobic interactions might occur between residues in the specificitydetermining region of the bacteriocin and a membrane component, thereby rendering the membrane more susceptible to permeabilization (56). Alternatively, membrane destabilization occurring due to increased bacteriocin binding to the cell surface and a higher concentration of bacteriocin molecules in the vicinity of the membrane as a result, appears less likely. Kaiser and Montville (15) hypothesized that, following hydrophobic interactions, the bacteriocin may be reoriented into a more energetically favorable orientation, which could simply be its insertion into the membrane followed by aggregation. As far as insertion of bacteriocins in the membrane bilayer is concerned, models exist for the lantibiotic nisin in various in vitro lipid bilayers and vesicles (30, 53), while for class IIa bacteriocins, it was only recently that a number of models have been proposed suggesting their possible orientations in lipid bilayers (12, 25, 29, 57, 58). Class IIa bacteriocins are in fact believed to insert into the target membrane via their hydrophobic and/or amphiphilic C-terminal domain, and aggregate to form water-filled pores (12, 25, 54). Class-lla-bacFactors affecting bacteriocin activity teriocin-induced effluxes and depletions of intracellular solutes in sensitive cells, leading to cell death, have been

VOL.87, 1999 shown to occur in concentration-dependent and timedependent manners (15, 47). Lethal activity of class IIa bacteriocins is further influenced by several other factors related either to the target cell or to the medium. In a recent investigation, Chen et al. (55) have confirmed bacteriocin-lipid interactions for pediocin PA1 and the absence of the requirement of membrane protein receptors. In parallel, it has been shown that the lipid composition of the target membrane is a determining factor in modulating the pediocin PA-l action, particularly the affinity of this bacteriocin for lipid vesicles increases with the increase in their anionic lipid content (55). The binding affinity of class IIa bacteriocins to target membranes has been found to be affected by pH. For instance, decreasing the pH from 7.5 to 6.0 has been shown to improve both membrane binding and permeabilization abilities of pediocin PA-I (29). Moreover, pore formation by bavaricin MN is optimal at pH 6.0 and less efficient at other pH values (15). Due to their relatively high content of basic residues, class IIa bacteriocins, as other bacteriocins, are peptides with a positive net charge under physiological conditions, which presumably allow them to adhere to negatively charged phospholipid head groups. Therefore, altering the charge properties of either the bacteriocin, by changing the medium pH, or the membrane, by changing its lipid composition, obviously influences this adhesion by affecting the dissociation constant of peptidelipid interactions. Both kinds of alterations further contirm the involvement of electrostatic binding in the interactions of class IIa bacteriocins with target membranes. Class IIa bacteriocins are generally opposed to nisin in the sense that they interact with the cytoplasmic membranes of sensitive cells regardless of their degree of prior energization, suggesting that the loss of permeability of the cytoplasmic membrane occurs in a voltageindependent manner (3, 43, 44, 48), while nisin acts in a membrane-potential-dependent manner (5, 53). However, recent investigations have shown, on the one hand, that dependency of nisin on A$ varies depending on the experimental system used: while a threshold level of A$ is required for activity in Listeria cells and black lipid membranes (5, 30, 43, 53, 59), nisin can, however, display activity on lipid vesicles and sensitive lactococcal cells in the absence of A$, though the presence of A$ increases its membrane permeabilization ability (5, 30, 43, 53). On the other hand, the antimicrobial activity of the class IIa bacteriocins, pediocin PA-l and bavaricin MN, has also been shown to be enhanced by A# (66% and 88% increase, respectively), although it is not fully dependent on it (15, 29). It has been suggested that these bacteriocins would function in an energy-enhanced manner (15). The presence of A$ is speculated to increase the size or the number of poration complexes and/or help in reorienting membrane-bound bacteriocin molecules into a more energetically favorable orientation, thereby promoting their insertion (15, 29). Spectrum of activity The bactericidal activity of class IIa bacteriocins seems to be targeting primarily the Listeria strains, since all these peptides show antilisterial activity, but it is also commonly directed against several other gram-positive bacteria. In fact, in addition to all species of Listeria, species belonging to the following genera have been reported to be sensitive to class IIa bacteriocins: Lactobacillus, Leuconostoc, Pediococcus,

CLASS IIa BACTERIOCINS OF LACTIC ACID BACTERIA

709

Lactococcus, Carnobacterium, Enterococcus, Micrococcus, Staphylococcus, Streptococcus, Clostridium, Bacillus and Brochothrix (2, 3, 9, 11, 12, 19, 23-25, 60). Furthermore, some class IIa bacteriocins have been shown to prevent the outgrowth of spores and vegetative cells of Clostridium spp. (19). Yet the extent of sensitivity to class IIa bacteriocins varies from species to species and from strain to strain. In other respects, it is noteworthy that class IIa bacteriocins display overall narrow inhibitory spectra, as compared to bacteriocins of other groups, such as nisin. Despite large regions of sequence identity and close homology in the N-terminus, class IIa bacteriocins have been reported to display considerable disparity in the spectra of antimicrobial activity (11, 23-25). In fact, the only apparent common characteristic of class IIa bacteriocins regarding their inhibitory spectra is their vigorous bactericidal action against strains of the genus Listeria. Therefore, the determination of factors accounting for the observed disparity would require detailed knowledge of various structural features of class IIa bacteriocins, such as their three-dimensional structures, which would have to be compared with their antibacterial profiles. However, studies which have matched antibacterial spectra generally used a maximum of four class IIa bacteriocins, which does not integrate sufficiently variable structural aspects, and/or a limited number of indicator strains. On the other hand, limited published data are available for class IIa bacteriocins that can be reasonably matched, due to the fact that the indicator species and strains and/or the methods used often differ from one study to another. So far, only the general features of the mode of action of class IIa bacteriocins, described above, have been identified, while factors underlying the specificity of action of these peptides are not yet clarified. Nevertheless, based on observations of antibacterial spectra, several interesting ideas have been proposed that can help structure-function studies to move forward. For instance, pediocin PA-l and enterocin A have been shown to exhibit spectra of activity broader than those of sakacin P and curvacin A, which has been ascribed to the presence of an extra disulfide bond in the two former bacteriocins (23). It has also been speculated that bacteriocins with fewer amino acid residues would tend to have a relatively broader antibacterial spectrum than those with larger numbers of residues (3). Furthermore, bacteriocins with slight differences in their structures, like mesentericin Y 105 and leucocin A, naturally display antibacterial spectra with insignificant differences or no difference at all (22, 25, 28). Surprisingly, class IIa bacteriocins from the same bacterium: carnobacteriocin BMl and B2 produced by C. piscicola LV 17B, and piscicocin Vla and Vlb produced by C. piscicola Vl, also exhibit spectra with only slight differences or no difference at all, although their respective structures are less related as compared to other members of the group (8, 12). Moreover, it has been recently shown that very few discrepancies in activity exist between four class IIa bacteriocins (pediocin PAl/AcH, enterocin A, curvacin A and sakacin P) when acting against strains within the genus Listeria, while within the LAB genera, results were far less consistent, as they varied considerably within each genus and species (23). In this regard, both specificity and potency of the activity of class IIa bacteriocins are apparently influenced by the lipid composition of target membranes which

710

ENNAHAR

ET AL.

J B~oscr. BIOENG.,

would determine the occurrence and the degree of interactions with bacteriocin molecules (15, 29, 46). Therefore, the comparable activities of class IIa bacteriocins against strains of Listeriu could be due to the similarity in membrane lipid composition within this genus. APPLICATIONS

IN FOOD SYSTEMS

Food preservation and safety In recent years, the food industry has evolved into centralized food processing operations and has decreased the number of people involved in food production. Furthermore, lifestyle changes relating to food preparation and consumption have increased the potential for the mishandling of foods during the various stages of processing, storing, distributing, retailing, and preparing. Modern developments have increased the life expectancy of humans, leading to an increasing proportion of immunocompromised individuals, who are more susceptible to illnesses caused by microorganisms, some of which are resistant to severe storage conditions such as refrigeration. At the same time, there has been an increasing demand among consumers for minimally processed foods that are not only convenient to use and available throughout the year, but also nutritious and economical. Food processors have met this demand by developing a wide variety of refrigerated foods with extended shelf life. These foods, however, by their very nature, require stringent conditions related to good manufacturing practices (GMPs), sanitation, hygiene, product formulation, storage temperature, and duration of refrigerated storage, therefore, presenting challenges to ensure high quality and minimal risk for foodborne illness. Pathogenic bacteria Among the foodborne pathogens that can grow at refrigeration temperatures (including nonproteolytic strains of Clostria’ium botulinum and some strains of Bacillus cereus), Ls. monocytogenes has had a tremendous impact on the public, by causing major listeriosis outbreaks worldwide. Recent outbreaks in North America and Europe, with an overall mortality rate of approximately 30% (61), have been traced to coleslaw, soft Mexican-style cheese, Vacherin Mont d’Or cheese, Brie cheese, rillettes (potted mince), pork tongue, pate, and milk (6, 62). Individuals with compromised immune systems, e.g., newborns, the elderly, and people suffering from the acquired immunodeficiency syndrome, are most susceptible to listeriosis, but healthy individuals may also acquire listeriosis (6). Ls. monocytogenes is ubiquitous in the environment, as it has been isolated from soil, silage and food-processing environments, and is basically capable of contaminating all raw animal products (6, 7, 63). In order to reduce/control the incidence of this pathogen in the food supply, regulatory authorities in many countries have set a Uzero tolerance” for several food products, which imposes severe processing conditions on food companies. Spoilage bacteria With sufficient time at refrigeration temperatures, several types of psychrotrophic/ psychrophilic bacteria may grow to levels sufficient to cause food spoilage. The microorganisms of primary concern in extended shelf life of refrigerated foods are Brochothrix thermosphacta, which has been recovered from vacuum-packaged beef, pork, lamb, and heatprocessed cured meats such as sliced cooked ham and corned beef (64), and LAB, which can spoil a variety of foods, including milk and milk products, meats, vegeta-

bles, fruit juices, sugary products, alcoholic beverages, and products preserved with vinegar (65). The extent of spoilage, which involves the development of sliminess and production of off-odors and off-flavors (66), varies with the fermentative type of LAB, product pH and oxygen permeability of the packaging material (1). Consequently, awareness of the importance of food preservation in extending food shelf life and wholesomeness and reducing potential health hazards has been growing, which is particularly evidenced by recent major actions taken by health and regulatory authorities in many countries, including new regulations and inspections that are based on the principles of the hazard analysis critical control point (HACCP) system. This is a systematic approach to the identification, evaluation, and control of food safety hazards, from raw material production and procurement to distribution and consumption of the finished product (1). Control measures Several types of control methods have been considered in preventing or minimizing microbial contamination of food and inhibiting the growth of or destroying microbial contaminants. Elementary measures are preventive ones and consist of GMPs, sanitation, and hygiene, which are necessary prerequisites for implementing an effective HACCP system, thus enabling the highest level of food safety assurance possible. However, since GMPs are often insufficient for preventing potential microbiological hazards, curative measures and processes are usually used. These consist of the following: (i) destroying microorganisms through heat (i.e., pasteurization, canning) or irradiation and (ii) inhibiting the growth of microorganisms through environmental control (e.g., refrigerating, freezing, drying and packaging), by adding chemical antimicrobials, or by adding desirable microorganisms in fermented foods (e.g., cheeses and pickles) to compete with undesirable microorganisms. A combination of several antimicrobial measures at subinhibitory concentrations is often used to achieve an effective control of microorganisms in refrigerated foods, which is referred to as the barrier/hurdle concept (67). Common hurdles include physical elements, such as refrigeration, modified atmosphere packaging and heat treatment, and physicochemical factors, such as natural product composition, water activity, pH, and added chemicals. When used together, hurdles interact, sometimes synergistically, enabling the use of lower intensities of each factor than would be necessary if each were used alone (1). This approach is very useful in food preservation because factors or antimicrobials should be applied at levels or concentrations having no adverse effects on product quality and safety. In recent years, because of consumer resistance to highly processed foods, various nonthermal preservation and shelf-life-extending methods have been identified and are being applied to the food industry. These include application of aseptic processing, hydrostatic pressure, electroporation, and LAB bacteriocins. Bacteriocins as food preservatives Preventive measures consisting of GMPs, sanitation, and hygiene have significantly contributed to the improvement of food safety in recent years. The use of bacteriocins is among the new approaches which may further contribute to reduce risks of foodborne disease outbreaks and increase food quality. Their use to inhibit pathogenic and spoilage organisms in food addresses a real need in the

VOL.

87, 1999

sense that they may preserve highly perishable foods, like meat and dairy products, whose contamination and deterioration are difficult to contain only by means of GMPs, sanitation, and hygiene. There is also potential for the use of bacteriocins to preserve foods intended for consumption or produced in developing countries, which have increasing needs for food but lack adequate refrigeration and distribution systems. However, although they act as strong bactericidal agents against sensitive microorganisms in culture media, bacteriocins in foods generally exhibit moderate antimicrobial activity (1 to 3 log cycles in Ls. monocytogenes populations (7)) followed by microbial growth. This has been ascribed to the inability of the bacteriocin to find all cell microenvironments within the food and/or its inactivation, which may be due to physical constraints posed by the relatively large size of the peptide and/or biochemical reactions with food components due to the proteinaceous nature of bacteriocins, respectively (7). Development of resistance in target cells is also often incriminated for the moderate efficiency of bacteriocins (7). Therefore, bacteriocins are not meant to be used as a primary means of food preservation, but they would appropriately be integrated in multihurdle preservation systems. For a bacteriocin to be useful as a natural food preservative, it must function in a food system. A number of bacteriocins exist that could be used immediately in foods, but few actually have been utilized, mainly because their effectiveness in foods has not been thoroughly investigated. For more efficient and extensive use of bacteriocins in food, application studies have to be carried out that may be very complex and include a number of factors, which can be evaluated in model systems or in actual foods. Particularly, there is a need to determine the types and number of microorganisms to be controlled, the bacteriocin efficacy in vitro and in food systems, its interactions with food components and other preservative systems, its mechanistic action, its application methods, and its possible sensory repercussions on food. In fact, although bacteriocins are generally effective at low concentrations, they may have an impact on sensory characteristics of fermented foods by affecting its microbial balance. Antimicrobial activity of bacteriotins often depends on the strain of target bacterium and the number of microorganisms present. The bacteriocin selected should not contribute to the development of resistant strains nor alter the environment of the food in such a way that growth of another pathogen is selected. Important intrinsic and extrinsic factors or variables associated with bacteriocin application to a food that need to be addressed during in vitro testing include temperature, atmosphere, pH, oxidation-reduction potential, and water activity. Another major factor that should be evaluated before food applications of bacteriocins is their isolation, purification and economical production, which can be complex, inefficient, and expensive. Bacteriocins may be produced in or used commercially as additives to the foods in which they are normally found or to other foods requiring preservation. In this regard, the fact that bacteriocins are naturally occurring antimicrobials is of great interest for the food industry, which has a long-standing interest in “natural” food preservation systems. Indeed, the negative perception of consumers towards industrially synthesized food additives has prompted food processors to label their prod-

CLASS

IIa BACTERIOCINS

OF LACTIC

ACID

BACTERIA

711

ucts as containing only those ingredients that consumers consider “natural”. Nevertheless, the highly purified bacteriocins, even though they come from natural sources, would need to be approved as food additives for use as food preservatives. This would involve expensive and time-consuming toxicological testing. In addition, they likely would have to be labeled as chemical additives, which would defeat the purpose of using bacteriocins as natural compounds. As a consequence, bacteriocins are most often recommended for use in fermented meat and dairy products through the producer bacterium, so that the dairy food has its own built-in food preservation system; one that does not require any regulatory approvals or label declarations. Among the many LAB bacteriocins discovered to date, there are only a few which were subjected to food trials, and even fewer applications. So far, the lantibiotic nisin, produced by certain strains of L. lactis subsp. lactis, remains the exclusive bacteriocin to be approved for food use and commercialized in over 50 countries where it is recognized as a safe food preservative (68). Established applications of nisin vary from one country to another and generally include control of: spoilage, particularly cheese blowing by Clostridium spp. of processed cheese; growth and toxin production by Clostridium botulinum in cheese spreads (69); spoilage of pasteurized milks and dairy desserts; and spoilage of canned vegetables due to heat-resistant spore-forming bacteria (68). More recently, in the US, generally recognized as safe (GRAS) petitions have been filled for the use of nisin in liquid whole eggs, sauces and salad dressings (70, 71). Moreover, Australia allowed the use of nisin to prevent the growth of Bacillus cereus in high moisture bakery products (68, 72). In addition numerous studies have demonstrated that other possibilities exist for application of nisin and nisin-producing cultures in controlling the growth of Ls. monocytogenes and Clostridium spp. in various dairy and meat products and in controlling spoilage LAB in alcoholic beverages (7, 68). Successful applications of nisin have spurred a growing interest in the potential practical use of bacteriocins in a wide variety of food products and motivated renewed attempts to characterize the complete repertoire of bacteriocins in order to select peptides with high activity and/or targeting a specific undesirable organism. Interest in class IIa bacteriocins Class IIa bacteriotins, due to their relatively narrow spectrum of activity, are of great interest, particularly in achieving specific elimination of an undesirable bacterium in fermented foods, while not affecting the useful microflora. In fact, compared to class IIa bacteriocins, nisin and other broadspectrum bacteriocins may be antagonistic towards cheese starter, lactococci, as well as other nonstarter LAB important for flavor development. In addition, studies using model food systems demonstrate that class IIa bacteriocins display an exceptionally high activity against Ls. monocytogenes and are more efficient at killing this pathogen in meat products, where nisin is ineffective (10, 30, 73-76). Due to this association of a narrow inhibitory spectrum and high antilisterial activity, class IIa bacteriocins are perhaps the most promising of the tested LAB bacteriocins for use in fermented foods that are prone to Listeria contaminations. Table 2 summarizes the studies on potential applications of class IIa bacteriocins as purified substances or through their producing strains used either as starter

712

ENNAHAR

ET AL.

J. Brosc~. BIOENC;.,

cultures for food fermentation or solely as biocontrol microorganisms or adjunct cultures. It appears that the majority of these studies have dealt with pediocin PA-

Leuconostoc gelidum UAL187,

l/AcH,

Lb. sake.

trated

whose antilisterial efficiency has been well illusin various

foods

and food

systems (Table

in Ls. monocytogenes

populations

(77-79).

Similar-

ly, the use of the sakacin-A producer Lb. sake Lb706 resulted in a l-log reduction of the pathogen in freshly minced adjunct

meat (80). On the other culture, pediocin-producing

hand,

when

respects,

a study

by Leisner TABLE

Purpose/Bacteriocin Inhibition

of Listeria

Pediocin PA-l/AcH

2.

tion as compared to an isogenic starter), while achieving normal

used as an

P. acidilactici failed to provide any protection against Lx monocytogenes in frankfurters at 4 and 15°C and in wieners at 4”C, which has been ascribed to the inability of the test strain to grow sufficiently at low temperatures, thereby, unable to produce inhibitory levels of bacteriocin (81, 82). At a higher temperature (25”(Z), however, antilisterial activity could be observed (82) (Table 2). The bavaricin-MN producer Lb. sake MN, used as an adjunct culture in beef cubes, has also been shown to display antilisterial activity closely dependent on temperature (83). In other et al. (84) has shown

that

(Table 2). In addition, Ennahar et al. (86) have shown that the natural pediocin PA-l/AcH-producing Lb. plantarum WHE92, inoculated as a surface adjunct culture during efficient

ripening at killing

of

Munster

cheese,

was

impressively

4log-cycle reduction as compared to bacteriocin-negative controls),

Ls. monocytogenes (approximately

without

Lb. pfantarum

altering

the cheese ripening

process.

WHE92, as well as genetically modified

lactococcal pediocin-producers, provide a great opportunity for efficient application of the beneficial properties of pediocin PA-l/AcH to the dairy industry compared to pediococci which are poorly adapted to colonizing dairy

Producer/bacteriocin preparation

Food/Food

system

Effect

Ref. -

monocytogenes Pediococcus

acidilactici

as a

Dry sausage Summer sausage Turkey summer sausage

Genetically transformed Lactococcus lactis MM217 as a starter culture coinoculated with the pathogen

Cheddar cheese

Pediococcus acidilactici as an adjunct culture coinoculated with the pathogen

Frankfurters Vacuum-packaged

Lactobacillus

plantarum

wieners

Munster cheese

WHE92 as an adjunct culture Purified bacteriocin

Food slurries sake MN as an adjunct culture coinoculated with the pathogen Lactobacillussake Lb706 as a starter culture Lactobacillus

Up to 1.6-log-cycle reduction Reduction of 1.2-log-cycle higher than in batteriocin-negative controls Reduction of 2.5log-cycle higher than in batteriocin-negative controls Reduction by 3 log cycles at 8”C, as compared to bacteriocin-negative controls at the end of ripening

(78) (77)

No effect at 4°C or 15°C

(81)

Reduction by 2.7 log cycles at 25°C but no effect at 4°C Up to 4-log-cycle reduction at 15”C, as compared to bacteriocin-negative controls at the end of ripening

(82)

(79) (85)

(86)

Reduction of attached cells by ca. 2 log cycles (75) Up to l-log-cycle reduction during heating at (7) 63°C Reduction by 1 log cycle and 3-log-cycle (88) difference compared to controls after 7 d at 4°C Increase in bacteriocin activity with Tween 80 (87) or liposome-encapsulated bacteriocin

Meat pieces Liquid whole egg Turkey slurries

Sakacin A

bacteriocin-negative cheese making (85)

Food preservation and safety mediated by class IIa bacteriocins and their producers

starter culture

Bavaricin MN

A producer,

As far as dairy products are concerned, cheddar cheese has been prepared with L. factis MM217, a starter culture genetically transformed for pediocin PA-l/AcH production, coinoculated with Ls. monocytogenes. The pediocin-producing starter significantly reduced the viability of the pathogen (approximately 3-log-cycle reduc-

2).

The inoculation of pediocin-producing P. acidilacfici as a starter culture in various types of fermented sausages provided between 1.2 and 2.5-log-cycle reduction

the leucocin

could be efficiently used as a barrier culture against the spoilage of vacuum-packaged beef by sulfide-producing

Beef cubes

At least l-log-cycle reduction at 4°C but no reduction at 10°C

(83)

Fresh minced meat

Reduction by ca. l-log-cycle as compared to bacteriocin-negative controls

(80)

Delay of the spoilage onset by 5 d at 2°C as compared to bacteriocin-negative controls

(84)

Spoilage control Leucocin A

Leuconostoc

geiidum

UAL187 as an adjunct culture coinoculated with a sulfide-producing Lactobacillus sake

Vacuum-packaged

beef

VOL. 87, 1999

products and are therefore less suitable for controlling Ls. monocytogenes in these foods. Studies indicate that other food products, particularly sausages, are capable of acquiring the same antilisterial protection by Lb. plantarum WHE92, a strain that is now being included in various dairy processes in France with particular success in commercial soft cheeses. Purified class IIa bacteriocins have also been examined for potential applications in foods and food systems, particularly pediocin PA-l/AcH has been shown to inhibit the growth of Ls. monocytogenes in liquid whole egg, meat and food slurries (7, 75, 87, 88). The use of class-IIa-bacteriocin-producing LAB as starters or adjunct cultures, as well as the use of the purified peptides in food products are still in experimental stages. In fact, although research in the area of molecular characterization of class IIa bacteriocins and their antimicrobial activity in culture media is very intense, studies dealing with bacteriocin efficiency in actual food systems remain quite limited. However, it is clearly apparent that class IIa bacteriocins may have future applications in improving the safety and the quality of a wide variety of food products, particularly in regards to extending the shelf life of meat and dairy products. This would certainly encourage further studies and developments towards applications of class IIa bacteriocins as food preservatives. CONCLUSION AND FUTURE PERSPECTIVES Class IIa bacteriocins are apparently abundant antimicrobials from LAB, which are highly effective in controlling the growth of undesirable microorganisms and attractive as food preservatives. The application of several class IIa bacteriocins as food preservatives is still in experimental stages and may have future applications in enhancing the safety and extending the shelf life of many inherently perishable foods. Pediocin PA-l is a representative example of potential applications of these peptides as a food preservatives. Its use in meats, cheeses and salads is actually covered by several US and European patents (Gonzalez et al., European Patent 88101624, 1988; Vandenbergh et al., European Patent 89101125.6, 1989; Bourdreaux et al., United States Patent 5137319, 1992; Werner-Aoude et al., French Patent 9411188, 1994). The need for expanded use of class IIa bacteriotins is obvious, especially in light of consumer demands for minimally processed, safe foods of adequate shelflife and convenience, and the global need for increasing the supply of food. Consequently, class IIa bacteriocins are believed to be the next in line if more bacteriocins are to be approved as food additives in the future. For the use of bacteriocins to increase, however, research, commercial development, and economic production must occur on an adequate scale; multihurdle preservative systems must be developed; technology transfer must occur. As far as research is concerned, the specific future challenges are discussed bellow. Heterologous expression, overexpression and multiexpression The limited efficiency of bacteriocin-producing cultures in fermented foods may be ascribed to various factors such as low production, regulatory systems, genetic instability, inactivation, and resistance development among target bacteria. A recent trend in bacteriotin research, which consists of heterologous expression of bacteriocins among various LAB strains, offers an

CLASS IIa BACTERIOCINS

OF LACTIC

ACID BACTERIA

713

excellent tool that may help overcome such obstacles. In particular, cloning and expression of class-IIa-bacteriotin genes in new hosts have allowed constitutive production and even overexpression of bacteriocins, therefore overcoming bacteriocin regulation systems and the related low-production problems (28, 89-92). Also, various food-grade LAB strains can be selected, based on their characteristics relevant to specific food systems, for use as hosts for defined bacteriocins of interest, therefore, yielding bacteriocin-producing LAB strains that are adapted to each type of food, which may help further overcome colonization and bacteriocin production problems (85). Moreover, the use of the heterologous expression of bacteriocins to develop LAB producing multiple bacteriocins, each one having its own specific range of target bacteria, may prove of great interest in enhancing the antimicrobial efficiency of LAB in food, by yielding bacteriocin-producing strains active against a broad range of undesirable organisms and possibly reducing the risk of bacteriocin-resistance development among target bacteria. Therefore, heterologous expression creates interesting possibilities for further development and extension of bacteriocin applications as preservatives in various food industries. Engineering of bacterlocins The engineering of bacteriocins, through genetic or chemical modifications, offers the possibility of developing new biologically important peptides with improved activity and stability. It is generally reported that sequence modifications, including single-residue substitutions, of class IIa bacteriocins result in peptides with diminished inhibitory activity compared to the native bacteriocins (24, 25, 29, 30, 93). However, a recent study by Miller et al. (94) showed a significant increase in pediocin PA-l activity upon substitution of a Glu residue for Lys-11, which suggests possible interesting future developments in the field of bacteriocin engineering. Multihurdle concept Obviously, food preservation can be enhanced not only by a class IIa bacteriocin with an individual antimicrobial effect, but, more interestingly, by combination of bacteriocins in order to enhance the overall effectiveness against target organisms. Class IIa bacteriocins are particularly suitable for use in multihurdle food preservation systems by interactions with multiple bacteriocins resulting in additive or synergistic effects that can better prevent or delay undesirable microbial activity. In fact, although they are all antilisterial compounds, largely vary in their spectra of activities, which allow many possibilities of combinations of class IIa bacteriocins for Listeria elimination procedures adapted to each food product. Combinations of class IIa bacteriocins with bacteriocins from other classes may also be useful for targeting multiple undesirable bacteria. Finally, as discussed above, for optimal effectiveness against foodborne pathogenic and spoilage bacteria, bacteriocins, including class IIa, have to be utilized as a part of a general multihurdle food preservation system, one which involves other antimicrobial factors, ACKNOWLEDGMENT

Financial support for research in the authors’ laboratory and the preparation of this manuscript, as well as a fellowship for S. E. were kindly provided by the Japan Society for the Promotion of Science.

714

ENNAHAR

ET AL.

J. BIOSCI. BIOENG..

REFERENCES 1. Martb, E. H.: Extended shelf life refrigerated foods: microbiological quality and safety. Food Technol., 52, 57-62 (1998). 2. Klaenbammer, T. R.: Genetics of bacteriocins produced by lactic acid bacteria. FEMS Microbial. Rev., 12, 39-86 (1993). 3. Jack, R. W., Tagg, J. R., and Ray, B.: Bacteriocins of grampositive bacteria. Microbial. Rev., 59, 171-200 (1995). 4. Nes, I. F., Diep, D. B., Hgvarstein, L. S., Brurberg, M. B., Eijsink, V., and Holo, H.: Biosynthesis of bacteriocins in lactic acid bacteria. Antonie Leeuwenhoek, 70, 113-128 (1996). 5. Sabl, H.-G. and Bierbaum, G.: Lantibiotics: biosynthesis and biological activities of uniquely modified peptides from grampositive bacteria. Annu. Rev. Microbial., 52, 41-79 (1998). 6. Farber, J. M. and Peterkin, P. I.: Listeria monocytogenes, a food-borne pathogen. Microbial. Rev., 55, 476-511 (1991). 7. Muriana, P.M.: Bacteriocins for control of Listeria spp. in food. .I. Food Prot., Supplement, 54-63 (1996). 8. Quad& L. E. N., Sailer, M., Roy, K. L., Vederas, J, C., and Stiles, M. E.: Chemical and genetic characterization of bacteriotins produced by Carnobacterium piscicola LVl7B. J. Biol. Chem., 269, 12204-12211 (1994). 9. Cintas, L. M., Casaus, P., Hgvarstein, L. S., Hermindez, P. E., and Nes, I. F.: Biochemical and genetic characterization of enterocin P, a novel set-dependent bacteriocin from Enterococcus faecium P13 with a broad antimicrobial spectrum. Appl. Environ. Microbial., 63, 4321-4330 (1997). 10. Jack, R. W., Wan, J., Gordon, J., Harmark, K., Davidson, D. E., Hillier, A. J., Wettenhall, R. E. H., Hickey, M. W., and Coventry, M. J.: Characterization of the chemical and antimicrobial properties of piscicolin 126, a bacteriocin produced by Carnobacterium piscicola JG126. Appl. Environ. Microbiol., 62, 2897-2903 (1996). 11. Aymerlcb, T., Holo, H., Hgvarstein, L. S., Hugas, M., Garrlga, M., and Na, I. F.: Biochemical and genetic characterization of enterocin A from Enterococcus faecium, a new antilisterial bacteriocin in the pediocin family of bacteriocins. Appl. Environ. Microbial., 62, 1676-1682 (1996). 12. Bhugaloo-Vial, P., Dousset, X., Mbtivier, A., Sorokine, O., Anglade, P., Boyaval, P., and Marion, D.: Purification and amino acid sequences of piscicocins Vla and Vlb, two class IIa bacteriocins secreted by Carnobacterium piscicola Vl that display significantly different levels of specific inhibitory activity. Appl. Environ. Microbial., 62, 4410-4416 (1996). 13. Hastings, J. W., Sailer, M., Johnson, K., Roy, K. L., Vederas, J.C., and Stiles, M.E.: Characterization of leucocin A-UAL 187 and cloning of the bacteriocin gene from Leuconostoc gelidum. J. Bacterial., 173, 7497-7500 (1991). 14. Henderson, J. T., Chopko, A. L., and van Wassenaar, P. D.: Purification and primary structure of pediocin PA-l produced by Pediococcus acidilactici PAC-1.0. Arch. Biochem. Biophys., 295,

5-12

(1992).

15. Kaiser, A. L. and Montville, T. J.: Purification of the bacteriotin bavaricin MN and characterization of its mode of action against Listeria monocytogenes Scott A cells and lipid vesicles. Appl. Environ. Microbial., 62, 4529-4535 (1996). mono16. Larsen, A. G. and Norrung, B.: Inhibition of Listeria cytogenes by bavaricin A, a bacteriocin produced by Lactobacillus bavaricus MI401. Lett. Appl. Microbial., 17, 132-

20.

21.

(1994). 22.

23.

24.

25.

26.

27.

28.

29.

HCcbard, Y., DCrljard, B., Letellier, F., and Cenatiempo, Y.: Characterization and purification of mesentericin Y 105, an anti-Listeria bacteriocin from Leuconostoc mesenteroides. J. Gen. Microbial., 138, 2725-2731 (1992). Eijsink, V. G. H., Skeie, M., Middelhoven, P. H., Brurberg, M. B., and Nes, I. F.: Comparative studies of class Ha bacteriotins of lactic acid bacteria. Appl. Environ. Microbial., 64, 3275-3281 (1998). Fimland, G., Blingsmo, 0. R., Sletten, K., Jung, G., Nes, I. F., and Nissen-Meyer, J.: New biologically active hybrid bacteriocins constructed by combining regions from various pediocin-like bacteriocins: the C-terminal region is important for determining specificity. Appl. Environ. Microbial., 62, 3313-3318 (1996). Fleury, Y., Abdel Dayem, M., Montagne, J. J., Chaboisseau, E,, Le Caer, J. P., Nicolas, P., and Delfour, A.: Covalent structure, synthesis, and structure-function studies of mesentericin Y 10537, a defensive peptide from gram-positive bacteria Leuconostoc mesenteroides. J. Biol. Chem., 271, 14421-14429 (1996). Abee, T.: Pore-forming bacteriocins of gram-positive bacteria and self-protection mechanisms of producer organisms. FEMS Microbial. Lett., 129, l-10 (1995). Sailer, M., Helms, G. L., Henkel, T., Niemczura, W. P., Stiles, M. E., and Vederas, J. C.: lsN- and L3C-labeled media from Anabaena sp. for universal isotopic labeling of bacteriotins: NMR resonance assignments of leucocin A from Leuconostoc gelidum and nisin A from Lactococcus lactis. Biochemistry, 32, 310-318 (1993). Fremaux, C., H&bard, Y., and Cenatiempo, Y.: Mesentericin Y105 gene clusters in Leuconostoc mesenteroides YIOS. Microbiology, 141, 1637-1645 (1995). Chen, Y., Sbapira, R., Eisenstein, M., and MontviBe, T. J.: Functional characterization of pediocin PA-l binding to liposomes in the absence of a protein receptor and its relationship to a predicted tertiary structure. Appl. Environ. Microbial., 63, 524-531

30.

31.

32.

134 (1993).

17. Tichaczek, P.S., Nissen-Meyer, J., Nes, I. F., Vogel, R. F., and Hammes. W. P.: Characterization of the bacteriocins curvacin A from Lactobacillus curvatus LTHl174 and sakacin P from L. sake LTH673. Syst. Appl. Microbial., 15, 460-468 (1992). 18. MCtivier, A., Pilet, M.-F., Dousset, X., Sorokine, O., Anglade, P., Zagorec, M., Pbud, J.-C., Marion, D., Cenatiempo, Y., and Fremaux, C.: Divercin V41, a new bacteriocin with two disulphide bonds produced by Carnobacterium divergens V41: primary structure and genomic organization. Microbiology, 144, 2837-2844 (1998). 19. Berm& M. H. J., Vanloo, B., Brasseur, R., Gorris, L. G. M., and Smid, E. J.: A novel bacteriocin with a YGNGV motif

from vegetable-associated Enterococcus mundtii: full characterization and interaction with target organisms. Biochim. Biophys. Acta, 1373, 47-58 (1998). Holck, A., Axelsson, L., Birkeland, S.-E., Aukrust, T., and Blom, H.: Purification and amino acid sequence of sakacin A, a bacteriocin from Lactobacillus sake Lb706. J. Gen. Microbiol., 138, 2715-2720 (1992). Tichaczek, P. S., Vogel, R. F., and Hammes, W. P.: Cloning and sequencing of sakP encoding, the bacteriocin produced by Lactobacillus sake LTH 673. Microbiology, 140, 361-367

33.

34.

35.

36.

(1997).

Montville, T. J. and Cben, Y.: Mechanistic action of pediocin and nisin: recent progress and unresolved questions. Appl. Microbial. Biotechnol., 50, 51 l-519 (1998). Fregeau Gallagher, N. L., Sailer, M., Niemczura, W. P., Nakasbima, T. T., Stiles, M. E., and Vederas, J. C.: Threedimensional structure of leucocin A in trifluoroethanol and dodecylphosphocholine micelles: spatial location of residues critical for biological activity in type IIa bacteriocins from lactic acid bacteria. Biochemistry, 36, 15062-15072 (1997). Casaus, P., Nilsen, T., Cintas, L. M., Nes, I. F., Herndndez, P. E., and Holo, H.: Enterocin B, a new bacteriocin from Enterococcus faecium T136 which can act synergistically with enterocin A. Microbiology, 143, 2287-2294 (1997). V. G. H.: PheroBrurberg, M. B., Nes, I. F., and Eijsink, mone-induced production of antimicrobial peptides in Lactobacillus. Mol. Microbial.. 26. 347-360 (19971. Bbuuia, A. K., Johnson,‘M.‘C., and R‘ay, B.: Direct detection of an antimicrobial peptide of Pediococcus acidilactici in sodium dodecyl sulfate-polyacrylamide gel electrophoresis. J. Ind. Microbial., 2, 319-322 (1987). Gonzalez, C. F. and Kunka, B. S.: Plasmid-associated bacteriocin production and sucrose fermentation in Pediococcus acidilactici. Appl. Environ. Microbial., 53, 2534-2538 (1987). Scbved, F., LaIazar, A., Henis, Y., and Juven, B. J.: Purification, partial characterization and plasmid-linkage of pediocin

VOL. 87, 1999 SJ-1, a bacteriocin produced by Pediococcus acidilactici. J. Appl. Bacterial., 74, 67-77 (1993). 37. Ray, S. K., Kim, W. J., Johnson, M. C., and Ray, B.: Conjugal transfer of a plasmid encoding bacteriocin production and immunity in Pediococcus acidilactici H. J. Appl. Bacteriol., 66, 393-399 (1989). 38. Bennik, M. H. J., Smid, E. J., and Gorrls, L. G. M.: Vegetable-associated Pediococcus parvulus produces pediocin PA-l. Appl. Environ. Microbial., 63, 2074-2076 (1997). 39. Ennahar, S., Aoude-Werner, D., Sorokine, O., van Dorsselear, A., Bringei, F., Hubert, J.-C., and Hasselmann, C.: Production of pediocin AcH by Lactobacifhs plantarum WHE 92 isolated from cheese. Appl. Environ. Microbial., 62, 4381-4387 (19%). 40. Tichaczek, P. S., Vogel, R. F., and Hammes, W. P.: Cloning and sequencing of curA encoding curvacin A, the bacteriocin produced by Lactobacillus curvatus LTH 1174. Arch. Microbiol., 160, 279-283 (1993). 41. Azelsson, L., Holck, A., Birkeland, S.-E., Aukrust, T., and Blom, H.: Cloning and nucleotide sequence of a gene from Lactobacillus sake Lb706 necessary for sakacin A production and immunity. Appl. Environ. Microbial., 59, 2868-2875 (1993). 42. Bhunia, A. K., Johnson, M. C., and Kaichayanand, N.: Mode of action of pediocin AcH from Pediococcus acidilactici H on sensitive bacterial strains. J. Appl. Bacterial., 70, 25-33 (1991). 43. Bruno, M. E. C. and Montviile, T. J.: Common mechanistic action of bacteriocin from lactic acid bacteria. Appl. Environ. Microbial., 59, 3003-3010 (1993). 44. Chikiadas, M. L., Garcia-Garcera, M. J., Driessen, A. J. M., Ledeboer, A. M., Nissen-Meyer, J., Nes, I. F., Abee, T., Konings, W. N., and Venema, G.: Pediocin PA-l, a bacteriocin from Pediococcus acidilactici PAC 1.0, forms hydrophilic pores in the cytoplasmic membrane of target cells. Appl. Environ. Microbial., 59, 3577-3584 (1993). 45. Montvilie, T. J. and Bruno, M. E. C.: Evidence that dissipation of proton motive force is a common mechanism of action for bacteriocins and other antimicrobial proteins. Int. J. Food Microbial., 24, 53-74 (1994). 46. Maftah, A., Renault, D., Vignoies, C., Hechard, Y., Bressoiiier, P., Rathtaud, M. H., Cenatiempo, Y., and Jtdien, R.: Membrane permeabilization of Listeria monocytogenes and mitochondria by the bacteriocin mesentericin YIOS. J. Bacterial., 175, 32323235 (1993). 47. Chen, Y. and Montvilie, T. J.: Efflux of ions and ATP depletion induced by pediocin PA-l are concomitant with cell death in Listeria monocytogenes Scott A. J. Appl. Bacterial., 79, 684-690 (1995). 48. Venema, K., Venema, G., and Kok, J.: Lactococcal bacteriotins: mode of action and immunity. Trends Microbial., 3, 299304 (1995). 49. Nicholls, D. G. and Ferguson, S. J.: Chemiosmotic energy transduction, p. I-20. In Nicholls, D. G. and Ferguson, S. J. (ed.), Bioenergetics 2. Academic Press, London (1992). 50. Abee, T., Rombouts, F. M., Hugenboltz, J., Guihard, G., and Leteiher, L.: Mode of action of nisin Z against Listeria monocytogenes Scott A grown at high and low temperatures. Appl. Environ. Microbial., 60, 1962-1968 (1994). 51. Winkowski, K., Bruno, M. E. C., and Montville, T. J.: Correlation of bioenergetic parameters with cell death in Listeria monocytogenes cells exposed to nisin. Appl. Environ. Microbiol., 60, 4186-4188 (1994). 52. Schved, F., Laiazar, A., Lindner, P., and Juven, B. J.: Iuteraction of the bacteriocin produced by Pediococcus acidilactici SJ-1 with the cell envelope of Lactobacillus spp. Lett. Appl. Microbial., 19, 281-283 (1994). 53. Mall, G. N., Roberts, G. C. K., Konings, W. N., and Driessen, A. J. M.: Mechanism of the lantibiotic-induced pore formation. Antonie Leeuwenhoek, 69, 185-195 (1996). 54. Chen, Y., Ludescher, R. D., and Montville, T. J.: Electrostatic interactions, but not the YGNGV consensus motif, govern the binding of pediocin PA-l and its fragments of phospholipid vesicles. Appl. Environ. Microbial., 63, 4770-4777 (1997).

CLASS IIa BACTERIOCINS

OF LACTIC ACID BACTERIA

715

55. Chen, Y., Ludescher, R. D., and Montville, T. J.: Influence of lipid composition on pediocin PA-1 binding to phospholipid vesicles. Appl. Environ. Microbial., 64, 3530-3532 (1998). 56. Fimiand, G., Jack, R., Jung, G., Nes, I. F., and Nissen-Meyer, J.: The bactericidal activity of pediocin PA-I is specifically inhibited by a 15-mer fragment that spans the bacteriocin from the center toward the C terminus. Appl. Environ. Microbial., 64, 5057-5060 (1998). 57. Franke, C. M., Leenhout, K. J., Haandrikman, A. J., Kok, J., Venema, G., and Veaema, K.: Topology of LcnD, a protein implicated in the transport of bacteriocins from Lactococcus la&. J. Bacterial., 176, 1766-1769 (1996). 58. Miiier, K. W., Schamber, R., Chen, Y., and Ray, B.: Production of active chimeric pediocin AcH in Escherichia coli in the absence of processing and secretion genes from the Pediococcuspap operon. Appl. Environ. Microbial., 64, 14-20 (1998). 59. Benz, R., Jung, G., and Sahl, H.-G.: Mechanism of channel formation by lantibiotics in black lipid membranes, p. 359372. In Jung, G. and Sahl, H.-G. (ed.), Nisin and novel lantibiotics. Escom, Leiden. The Netherlands (1991). 60. Larsen, A. G., Vogenskn, F. K., and Josephsen, J.: Antimicrobial activity of lactic acid bacteria isolated from sour doughs: purification and characterization of bavaricin A, a bacteriocin produced by Lactobacillus bavaricus MI401. J. Appl. Bacterial., 75, 113-122 (1993). 61. Griffiths, M. W.: Lhteria monocytogenes: its importance in the dairy industry. J. Sci. Food Agr., 47, 133-158 (1989). 62. McLauchiin, J.: The relationship between Listeria and listeriosis. Food Control, 7, 187-193 (1996). 63. Ryser, E. T. and Marth, E. H.: Ltiteria, listeriosis and food safety. Marcel Dekker, Inc., New York (1991). 64. Kraft, A.A.: Psychrotrophic bacteria in foods: disease and spoilage. CRC press, Boca Raton, Florida, USA (1992). 65. Sharpe, M. E. and Pettipher, G. L.: Food spoilage by lactic acid bacteria, p. 200. In Rose, A. H. (ed.), Food microbiology, vol. 8. Academic Press, New York (1983). 66. Jay, J. M.: Modern food microbiology, 5” ed. Chapman & Hall, New York (1992). 67. Leistner, L. and Gorrls, L. G. M.: Food preservation by hurdle technology. Trends Food Sci., 6, 41-46 (1995). 68. Delves-Broughton, J., Blackburn, P., Evan, R. J., and Hugenholtz, J.: Applications of the bacteriocin, nisin. Antonie Leeuwenhoek, 69, 193-202 (1996). 69. FDA: Nisin preparation: affirmation of GRAS status as a direct human food ingredient. Food and Drug Administration, Fed. Regist., 53, 11247-11251 (1988). 70. FDA: M. G. Waldbaum Co.: Filling of petition for affirmation of GRAS status. Food and Drug Administration, Fed. Regist., 59, 12582-12583 (1994). 71 FDA: Aplin & Barrett Ltd.: Filling of petition for alhrmation of GRAS status. Food and Drug Administration. Fed. Reqist., . 60, 64167 (1995). 72. Jenson, I., Baird, L., and Delves-Broughton, J.: The use of nisin as a preservative in crumpets. J. Food Prot., 57, 874-877 (1994). 73. Baccus-Taylor, G., Glass, K. A., and Luchansky, J. B.: Fate of Listeria monocytogenes and pediococcal starter cultures during the manufacture of chicken summer sausage. Poultry Sci., 72, 1772-1778 (1993). 74. Pucci, M. J., Vedamuthu, E. R., Kunka, B. S., and Vandenbergh, P. A.: Inhibition of Listeria monocytogenes by using bacteriocin PA-1 produced by Pediococcus acidilactici PAC 1.0. Appl. Environ. Microbial., 54, 2349-2353 (1988). 75. Nielsen, J. W., Dickson, J. S., and Grouse, J. D.: Use of bacteriocin produced by Pediococcus acidilactici to inhibit Listeria monocytogenes associated with fresh meat. Appl. Environ. Microbial., 56, 2142-2145 (1990). 76. Cintas, L. M., Casaus, P., Fernandez, M. F., and Hemandez, P. E.: Comparative antimicrobial activity of enterocin L50, pediocin PA-l, nisin A and lactocin S against spoilage and foodborne pathogenic bacteria. Food Microbial.. 15. 289-298 (1998). 77. Berry, E. D., Liewen, M. B., Mandigo, R. W., and Hutkins. ,A.

716

78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

ENNAHAR

ET AL.

R. W.: Inhibition of Listeria monocytogenes by bacteriocinproducing Pediococcus during the manufacture of fermented semi-dry sausage. J. Food Prot., 53, 194-197 (1990). Foegeding, P.M., Thomas, A. B., Pilkington, D. H., and Klaenhammer, T.R.: Enhanced control of Lisferia monocytogenes by in situ-produced pediocin during dry fermented sausage production. Appl. Environ. Microbial., 58, 884-890 (1992). Luchansky, J. B., Glass, K. A., Harsono, K. D., Degnan, A. J., Faith, N. G., Cauvin, B., Baccus-Taylor, G., Arihara, K., Bater, B., Maurer, A. J., and Gassens, R. G.: Genomic analysis of Pediococcus starter cultures used to control Listeria monocytogenes in turkey summer sausage. Appl. Environ. Microbial., 58, 3053-3059 (1992). Schillinger, U., Kaya, M., and Lticke, F.-K.: Behaviour of Listeria monocytogenes in meat and its control by a bacteriotin-producing strain of Lactobacillus sake. J. Appl. Bacterial., 70, 473-478 (1991). Berry, E. D., Hutkins, R, W., and Mandigo, R. W.: The use of bacteriocin-producing Pediococcus acidilactici to control postprocessing Listeria monocytogenes contamination of frankfurters. J. Food Prot., 54, 681-686 (1991). Degnan, A. J., Yousef, A. E., and Luchansky, J. B.: Use of Pediococcus acidilactici to control Listeria monocytogenes in temperature-abused vacuum-packaged Wieners. J. Food Prot., 55, 98-103 (1992). Winkoswki, K., Crandall, A. D., and Montville, T. J.: Inhibition of Listeria monocytogenes by Lactobacillus bavaricus MN in beef systems at refrigeration temperatures. Appl. Environ. Microbial., 59, 2552-2557 (1993). Leisner, J. J., Greer, G. G., and Stiles, M. E.: Control of beef spoilage by a sulfide-producing Lactobacillus sake strain with bacteriocinogenic Leuconostoc gelidum UALl87 during anaerobic storage at 2°C. Appl. Environ. Microbial., 62, 2610-2614 (1996). Buyoug, N., Kok, J., and Luchansky, 3. B.: Use of a genetically enhanced, pediocin-producing starter culture, Lactococcus lactis subsp. lactis MM217, to control Listeria monocytogenes in Cheddar cheese. Appl. Environ. Microbial., 64, 4842-4845 (1998). Ennahar, S., Assobhei, O., and Hasselmann, C.: Inhibition of Listeria monocytogenes in a smear-surface soft cheese by Lactobacillus plantarum WHE 92, a pediocin AcH producer. J. Food Prot., 61, 186-191 (1998). Degoan, A. J., Buyong, N., and Luchansky, J. B.: Antilisterial activity of pediocin AcH in model food systems in the presence of an emulsifier or encapsulated within liposomes. Int. J. Food Microbial., 18, 127-138 (1993). Schlyter, J. H., Glass, K. A., Loeffelholz, J., Degnan, A. J., and Luchausky, J. B.: The effects of diacetate with nitrite, lactate, or pediocin on the viability of Listeria monocytogenes in turkey slurries. Int. J. Food Microbial., 19, 271-281 (1993). McCormick, J. K., Worobo, R, W., and Stiles, M. E.: Expression of the antimicrobial peptide carnobacteriocin B2 by a

J BIOSCI. BIOENG..

90.

91

92.

93. 94.

95.

96. 97.

98.

99.

100.

101. 102.

signal peptide-dependent general secretory pathway. Appl Environ. Microbial., 62, 4095-4099 (1996). Biet, F., Berjeaud, J. M., Worobo, R. W., Cenatiempo, Y., and Fremaux, C.: Heterologous expression of the bacteriocin mesentericin Y105 using the dedicated transport system and the general secretion pathway. Microbiology, 144, 2845-2854 (1998). Horn, N., Martinez, M. I., Martinez, J. AI., Heruandez, P. E., Gasson, M. J., Rodriguez, J. M., and Dodd, H. M.: Production of pediocin PA-l by Lactococcus lactis using the lactococcin A secretory apparatus. Appl. Environ. Microbial., 64, 818-823 (1998). Chikindas, M. L., Venema, K., Ledeboer, A.M., Veuema, G., and Kok, J.: Expression of lactococcin A and pediocin PA-l in heterologous hosts. Lett. Appl. Microbial., 21, 183189 (1995). Quadri, L. E. N., Yan, L. Z., Stiles, M. E., and Vederas, J. C.: Effect of amino acid substitutions on the activity of carnobacteriocin B2. J. Biol. Chem., 272, 3384-3388 (1997). Miller, K. W., Schamber, R., Osmanagaoglu, O., and Ray, B.: Isolation and characterization of pediocin AcH chimeric protein mutants with altered bactericidal activity. Appl. Environ. Microbial., 64, 1997-2005 (1998). Motlagh, A. M., Bhunia, A. K., Szostek, F., Hansen, T. R., aud Ray, B.: Nucleotide, and amino acid sequence of papgene (pediocin AcH production) in Pediococcus acidilactici H. Lett. Appl. Microbial., 15, 45-48 (1992). Hastinas. J. W. and Stiles. M. E.: Antibiosis of Leuconostoc gelidum isolated from meat. J. Appl. Bacterial., 70, 127-134 (1991). Felix, J. V., Papathauasopoulos, M. A., Smith, A. A., and von Holy, A.: Characterization of leucocin B-Talla: a bacteriocin from Leuconostoc carnosum Talla isolated from meat. Curr. Microbial., 29, 207-212 (1994). Pilet, M.-F., Dousset, X., Barre, R., Novel, G., Desmazeaud, M., and Piard, J.-C.: Evidence of two bacteriocins produced by Carnobacterium piscicola and Carnobacterium divergens isolated from fish and active against Listeria monocytogenes. J. Food Prot., 58, 256-262 (1995). Holck, A. L., Axelsson, L., Htihne, K., and Kriickel, L.: Purification and cloning of sakacin 674, a bacteriocin from Lactobacillus sake Lb674. FEMS Microbial. Lett., 115, 143-150 (1994). Lewus, C. B., Kaiser, A., and Montville, T. J.: Inhibition of food-borne bacterial pathogens by bacteriocins from lactic acid bacteria isolated from meat. Appl. Environ. Microbial., 57, 1683-1688 (1991). Scbillinger, U. and Lticke, F.-K.: Antibacterial activity of Lactobacillus sake isolated from meat. Appl. Environ. Microbiol., 55, 1901-1906 (1989). Ahu, C. and Stiles, M. E.: Plasmid-associated bacteriocin production by a strain of Carnobacterium piscicola from meat. Appl. Environ. Microbial., 56, 2503-2510 (1990).