|Home | About | Journals | Submit | Contact Us | Français|
Lactococcin G and enterocin 1071 are two homologous two-peptide bacteriocins. Expression vectors containing the gene encoding the putative lactococcin G immunity protein (lagC) or the gene encoding the enterocin 1071 immunity protein (entI) were constructed and introduced into strains sensitive to one or both of the bacteriocins. Strains that were sensitive to lactococcin G became immune to lactococcin G when expressing the putative lactococcin G immunity protein, indicating that the lagC gene in fact encodes a protein involved in lactococcin G immunity. To determine which peptide or parts of the peptide(s) of each bacteriocin that are recognized by the cognate immunity protein, combinations of wild-type peptides and hybrid peptides from the two bacteriocins were assayed against strains expressing either of the two immunity proteins. The lactococcin G immunity protein rendered the enterococcus strain but not the lactococcus strains resistant to enterocin 1071, indicating that the functionality of the immunity protein depends on a cellular component. Moreover, regions important for recognition by the immunity protein were identified in both peptides (Lcn-α and Lcn-β) constituting lactococcin G. These regions include the N-terminal end of Lcn-α (residues 1 to 13) and the C-terminal part of Lcn-β (residues 14 to 24). According to a previously proposed structural model of lactococcin G, these regions will be positioned adjacent to each other in the transmembrane helix-helix structure, and the model thus accommodates the present results.
Lactic acid bacteria (LAB) produce ribosomally synthesized antimicrobial peptides, generally referred to as bacteriocins. There are two main classes of these bacteriocins (8, 22): the class I bacteriocins (often referred to as lantibiotics) that contain the modified amino acid residues lanthionine and/or β-methyllanthionine and the class II bacteriocins that lack modified residues (8). The class II bacteriocins are further divided into four subclasses, IIa, IIb, IIc, and IId (8). Class IIa contains the pediocin-like bacteriocins, which have very similar amino acid sequences, class IIc consists of the cyclic bacteriocins, and the one-peptide, noncyclic bacteriocins that show no sequence similarity to the pediocin-like bacteriocins are placed in class IId (8). The unmodified two-peptide bacteriocins are placed in class IIb. They are unique in that they consist of two different peptides, both of which must be present, in about equal amounts, to obtain optimal antimicrobial activity (25). More than 10 two-peptide bacteriocins have been isolated and characterized (see reference 25 for original references) since the first isolation of such a bacteriocin (lactococcin G) in 1992 (21). For the two-peptide bacteriocins that have been genetically characterized, the genes encoding the two bacteriocin peptides are always found next to each other in the same operon, along with the gene encoding the immunity protein that protects the bacteriocin producer from being killed by its own bacteriocin.
Lactococcin G is perhaps the best-characterized two-peptide bacteriocin (12, 18, 19, 21, 24, 26, 27). It consists of the 39-residue α peptide (termed Lcn-α) and the 35-residue β peptide (termed Lcn-β) (Fig. (Fig.1A).1A). Like all two-peptide bacteriocins whose mode of action has been studied, lactococcin G causes cell death by rendering the membranes of target cells permeable to various ions (18, 19). Nuclear magnetic resonance (NMR) and circular dichroism (CD) spectroscopy have revealed that the two lactococcin G peptides adopt mainly α-helical structures when they are individually exposed to membrane-like entities (12, 27). Based on the NMR structures and findings from site-directed mutagenesis studies, a structural model of lactococcin G has recently been proposed (23, 26, 27). In this model, the two complementary peptides form parallel helices that span the target cell membrane. The helix-helix segment consists of the N-terminal region of Lcn-α (from about Trp-3 to Gly-22) and the C-terminal region of Lcn-β (from about Tyr-13 to Trp-32). The model also proposes that the cationic C-terminal end (residues 35 to 39, R-K-K-K-H) of Lcn-α is unstructured and forced through the target cell membrane by the membrane potential, thereby positioning the C termini of the two peptides inside the target-cell (Fig. (Fig.2).2). The tryptophan-rich N-terminal end of Lcn-β is also proposed to be relatively unstructured and to position itself in the outer membrane interface, thus forcing the N termini of the two peptides to remain on the outer side of the target cell membrane and the helix-helix segment to transverse the membrane (Fig. (Fig.2).2). This proposed structure is presumably also valid for the two-peptide bacteriocins enterocin 1071 (4, 5, 11), enterocin C (17), and lactococcin Q (32), since their sequence similarities to lactococcin G (lactococcin G has about 88 and 57% sequence identity to lactococcin Q and enterocin 1071, respectively; enterocin 1071 and enterocin C are identical except for one residue) indicate that these four bacteriocins have similar three-dimensional structures.
The sequence of the lactococcin G operon (GenBank accession no. FJ938036) has been determined, and a gene (lagC) encoding the putative lactococcin G immunity protein has been identified downstream of the two genes encoding the two lactococcin G peptides. Downstream of the two genes encoding the two enterocin 1071 peptides, a gene (entI) encoding the enterocin 1071 immunity protein has been identified (4, 11). The putative lactococcin G immunity protein shows 38% amino acid sequence identity to the enterocin 1071 immunity protein (the sequence of which was obtained from Franz et al. ), and both proteins consist of 110 amino acid residues (Fig. (Fig.1A).1A). The aim of this study was to identify which peptides or which parts of the peptides of the two-peptide bacteriocins lactococcin G and enterocin 1071 are recognized by these immunity proteins. To achieve this, combinations of wild-type lactococcin G peptides, wild-type enterocin 1071 peptides, and hybrid lactococcin-enterocin peptides were assayed against sensitive strains that were transformed with an expression plasmid carrying either the lactococcin G or the enterocin 1071 immunity gene.
Escherichia coli DH5α cells were grown with vigorous agitation at 37°C in Luria-Bertani medium. For E. coli DH5α containing plasmid pLT100, pMG36e, pLcnGim, or pEnt1071im, erythromycin was added to a final concentration of 150 μg/ml for selection. For E. coli cells containing the plasmid pGEM-7Zf(+) (Promega) or pGEM-T Easy (Promega), ampicillin was added to a final concentration of 100 μg/ml. For growth on agar plates, the medium was solidified by the addition of 1.5% (wt/vol) agar.
Lactococcus lactis LMGT-2081 is a naturally lactococcin G-producing strain and was used to isolate template DNA for amplification of the lactococcin G immunity gene by PCR. The strain was grown without agitation at 30°C in GM17 medium (M17 medium [Oxoid] supplemented with 0.4% [wt/vol] glucose).
Lactobacillus sakei Lb790, containing the two-plasmid expression system used for production of different bacteriocin peptide variants, was grown without agitation at 30°C in MRS broth (Oxoid). The two plasmids pSAK20 and pLT100 contain markers of resistance to chloramphenicol and erythromycin, respectively. Consequently, these antibiotics were added to the growth medium, each at a final concentration of 10 μg/ml. For growth on agar plates, the medium was solidified with 1.5% (wt/vol) agar and the antibiotic concentrations were reduced to 2 μg/ml for erythromycin and 5 μg/ml for chloramphenicol.
The indicator organisms used in the activity assay were Lactococcus strain LMGT-2077, L. lactis MG1363, and Enterococcus faecalis LMGT-2333. The lactococcus strains were grown without agitation at 30°C in GM17 medium containing 0.1% (vol/vol) Tween 80, whereas E. faecalis LMGT-2333 was grown without agitation in MRS broth (Oxoid) at 30°C. When the indicator strains contained the plasmid pMG36e or the plasmid carrying the lactococcin G immunity gene (pLcnGim) or the enterocin 1071 immunity gene (pEnt1071im), erythromycin was added to the growth medium (2 μg/ml erythromycin for the lactococcus strains and 5 μg/ml erythromycin for the enterococcus strain) to ensure that the plasmids were kept in the cells.
To construct the plasmid containing the lactococcin G immunity gene, the putative immunity gene (lagC) was amplified by PCR using a PTC-200 Peltier thermal cycler (MJ Research), Taq DNA polymerase (Fermentas), and chromosomal DNA from the naturally lactococcin G-producing strain L. lactis LMGT-2081 as the template. The 50-μl PCR mixture contained about 50 ng of the template, 200 ng of each primer (Eurogentec), each deoxynucleoside triphosphate (Stratagene) at a final concentration of 0.1 mM, and 2.5 U of Taq DNA polymerase. The primers used in the PCR are listed in Table Table11 and were termed LcnGimR and LcnGimF. The immunity gene was amplified over 30 cycles of denaturation for 45 s at 95°C, annealing for 1 min at 48°C, and polymerization for 2 min at 72°C. The resulting PCR product was sequenced for verification, and the verified PCR product was cloned into pGEM-7Zf(+) (Promega) by using the BamHI and EcoRI restriction sites of both the PCR product and the plasmid. The lactococcin G immunity gene was then cloned into the E. coli/LAB shuttle vector pMG36e (31) by using the SacI and XbaI restriction sites of both the pGEM-7Zf(+) and pMG36e plasmids. The resulting plasmid was termed pLcnGim.
A pUC57 plasmid containing the enterocin 1071 immunity gene (entI, the sequence of which was obtained from Franz et al. ) was synthesized by GenScript Corporation (NJ) and was used as a template to amplify the enterocin 1071 immunity gene with Taq DNA polymerase (Fermentas). The PCR was performed under the same conditions employed for the amplification of the lactococcin G immunity gene. The primers (Eurogentec) used in the PCR are listed in Table Table11 and were termed Ent1071imR and Ent1071imF. The resulting PCR product was sequenced for verification, and the verified PCR product was cloned into a pGEM-T Easy vector (Promega) before it was cloned into pMG36e by using the SacI and XbaI restriction sites of the PCR product and the pMG36e plasmid. The resulting plasmid was termed pEnt1071im.
Competent E. coli DH5α cells were prepared using the CaCl2 method (28), and competent cells were transformed by heat shock at 42°C for 90 s. L. sakei Lb790/pSAK20 cells were made electrocompetent basically as described by Aukrust et al. (1). The cells were grown in MRS broth (Oxoid) supplemented with 2% (wt/vol) glycine and 10 μg/ml chloramphenicol to an optical density at 600 nm of 0.5 to 0.6. The cells were washed and transformed by electroporation using a Gene Pulser apparatus and a Pulse Controller unit (Bio-Rad Laboratories) as described in procedure 2 by Aukrust et al. (1). Electrocompetent Lactococcus strain LMGT-2077 and L. lactis MG1363 were prepared and transformed by electroporation as described by Holo and Nes (14). Electrocompetent E. faecalis LMGT-2333 was prepared and transformed by electroporation as described by Shepard and Gilmore (29), except that the cells were grown in SGM17 growth medium containing 6% (wt/vol) glycine.
Peptides were produced by a two-plasmid expression system using L. sakei Lb790/pSAK20 cells transformed with previously constructed plasmids (derived from pLT100) carrying the gene encoding either one of the wild-type peptides (Lcn-α, Lcn-β, Ent1071A, or Ent1071B) or one of the hybrid peptides (the Lcn-α-Ent1071A hybrid designated α[1-16]/A[14-39] or the Lcn-β-Ent1071B hybrid designated β[1-13]/B[11-35]) (24). The two-plasmid expression system is based on the production and secretion of the class IIa bacteriocin sakacin A (2, 3). The peptides were purified using a cation-exchange column, followed by reverse-phase chromatography, as described previously (24, 30). Depending on the peptide purified, between 100 and 500 μg of peptide per liter of culture was obtained. The purity of the peptides was determined to be >80% by analytical reverse-phase chromatography using a μRPC SC 2.1/10 C2/C18 column (GE Healthcare) in the SMART chromatography system (GE Healthcare).
The concentrations of the peptides were calculated by using UV absorption at 280 nm and the molar extinction coefficients calculated from the individual amino acid residues.
Combinations of the wild-type peptides and hybrid peptides were assayed for antimicrobial activity by using a microtiter plate assay system as described previously (21). The peptide combination (in twofold dilutions) and the indicator strain (Lactococcus strain LMGT-2077, L. lactis MG1363, or E. faecalis LMGT-2333) were added to a final volume of 200 μl in each well. Growth medium for the lactococcus strains was GM17 containing 0.1% (vol/vol) Tween 80, whereas E. faecalis LMGT-2333 was grown in MRS broth. Stationary-phase cultures of the lactococcus strains were diluted 1:50 before being added to the microtiter plates, and the plates were incubated for 5 h at 30°C. Likewise, stationary-phase cultures of E. faecalis LMGT-2333 were diluted 1:200 and incubated overnight at 30°C. Growth inhibition was measured spectrophotometrically at 600 nm with a microplate reader (Tecan), and the MIC was determined. The MIC was defined as the total amount of peptides (the sum of concentrations of both peptides in a 1:1 ratio) that inhibited growth by 50%. The degree of protection conferred by the two immunity proteins was evaluated by comparing the MICs obtained for an indicator strain expressing the immunity gene and the MICs obtained for the same strain containing the unmodified pMG36e plasmid.
The gene encoding the enterocin 1071 immunity protein (entI) has been shown previously to encode a protein that confers immunity toward enterocin 1071 in an enterocin 1071-sensitive strain (11). In the lactococcin G operon, the lagC gene encodes the putative lactococcin G immunity protein (13, 20). Based on the DNA sequence of the operon (GenBank accession no. FJ938036), PCR primers were designed to amplify the putative lactococcin G immunity gene from a natural lactococcin G producer (L. lactis LMGT-2081). The PCR product was cloned into the pMG36e expression vector and sequenced. The DNA sequence was identical to that of the putative lactococcin G immunity gene (lagC) identified by Nes et al. (GenBank accession no. FJ938036) (20). Bacterial strains sensitive to lactococcin G and/or enterocin 1071 were transformed with an expression plasmid carrying the gene encoding the putative lactococcin G immunity protein (the plasmid termed pLcnGim) or the enterocin 1071 immunity protein (the plasmid termed pEnt1071im). These strains were then assayed against different combinations of wild-type lactococcin G peptides, wild-type enterocin 1071 peptides, and hybrid peptides from lactococcin G and enterocin 1071 in order to determine which peptides and which parts of the peptides are involved in the recognition of the bacteriocins by the cognate immunity proteins. The degree of protection was evaluated by comparing the MICs obtained for an indicator strain expressing the immunity protein with the MICs obtained for the same strain containing the unmodified pMG36e plasmid.
Both immunity proteins conferred immunity toward the cognate bacteriocin (Table (Table2).2). For strains expressing the putative lactococcin G immunity protein, 2,000- to 14,000-fold increases (depending on the strain) (Table (Table2)2) in the concentrations of the two lactococcin G peptides (i.e., Lcn-α and Lcn-β) were required to obtain antimicrobial activity similar to that obtained for the strains not expressing the putative immunity protein, indicating that the lagC gene in fact encodes the protein involved in lactococcin G immunity. For strains expressing the enterocin 1071 immunity protein, 400- to 54,000-fold increases (depending on the strain) (Table (Table2)2) in the concentrations of the two enterocin 1071 peptides (i.e., Ent1071A and Ent1071B) were necessary to obtain the same antimicrobial effect obtained for the strains not expressing the immunity protein. Neither of the two immunity proteins protected the lactococcus strains against the two-peptide bacteriocin plantaricin EF, a bacteriocin which is unrelated to lactococcin G and enterocin 1071 (Table (Table2).2). The immunity proteins, however, provided some degree of cross-immunity to the homologous bacteriocins lactococcin G and enterocin 1071. The enterocin 1071 immunity protein protected the lactococcus strains against both enterocin 1071 and lactococcin G, while the lactococcin G immunity protein protected the enterococcus strain, but not the lactococcus strains, against enterocin 1071 (Table (Table22).
It has been speculated previously that the two-peptide bacteriocins recognize target cells through a receptor in the target cell membrane (26). Recently, it was shown that a component of the mannose phosphotransferase system (mannose-PTS) is the target cell receptor for several class II bacteriocins such as lactococcin A and the pediocin-like bacteriocins (9). It was also demonstrated that the bacteriocin-mannose-PTS complex is recognized directly by the cognate bacteriocin immunity protein but only when the bacteriocin is bound to the mannose-PTS (9). No such interactions between the mannose-PTS, bacteriocins, and the cognate immunity proteins has been detected for the two-peptide bacteriocins lactococcin G and plantaricin EF (9), suggesting that lactococcin G and other two-peptide bacteriocins may use another cellular component as a target receptor. The fact that the lactococcin G immunity protein confers immunity (46,000-fold protection) toward enterocin 1071 in the enterococcus strain but does not confer immunity to the same extent (only 2- to 40-fold protection, depending on the strain) in the lactococcus strains (Table (Table2)2) indicates that a cellular component is indeed required for the lactococcin G immunity protein to function. The lactococcin G immunity protein was clearly expressed in the lactococcus strains, since these strains were well protected against lactococcin G (with 2,000- to 14,000-fold protection, depending on the strain) (Table (Table22).
To identify which of the peptides or which parts of the peptides of lactococcin G and enterocin 1071 are recognized by the cognate immunity proteins, the antimicrobial activities of Lcn-α, Ent1071A, and an Lcn-α-Ent1071A hybrid peptide in combination with either Lcn-β, Ent1071B, or an Lcn-β-Ent1071B hybrid peptide were assayed for two strains (L. lactis MG1363 and Lactococcus strain LMGT-2077) expressing either the lactococcin G or the enterocin 1071 immunity protein (Table (Table3).3). The amino acid sequences of the two hybrid peptides are shown in Fig. Fig.1B.1B. The Lcn-α-Ent1071A hybrid peptide (termed α[1-16]/A[14-39]) consists of the N-terminal end of Lcn-α (residues 1 to 16) and the C-terminal end of Ent1071A (residues 14 to 39). Residues 14 to 16 in Lcn-α and Ent1071A are identical. The Lcn-β-Ent1071B hybrid peptide (termed β[1-13]/B[11-35]) consists of the N-terminal end of Lcn-β (residues 1 to 13) and the C-terminal end of Ent1071B (residues 11 to 35). Residues 11 to 13 in Lcn-β and Ent1071B are identical. Of the nine possible peptide combinations, two combinations, Lcn-α+Ent1071B and α[1-16]/A[14-39]+Ent1071B, could not be evaluated because their antimicrobial activities were too low (with MICs of 500 to 1,000 nM). Data for these two peptide combinations are consequently not included in Table Table33.
The lactococcin G immunity protein clearly protected both indicator strains efficiently against both the Lcn-α+β[1-13]/B[11-35] and the α[1-16]/A[14-39]+β[1-13]/B[11-35] combinations but did not protect the two strains against the Ent1071A+β[1-13]/B[11-35] combination (Table (Table3).3). There are consequently residues in the N-terminal region (up to and including residue 13) of Lcn-α that are important for recognition of lactococcin G by the immunity protein, and these residues are absent in the corresponding N-terminal region of Ent1071A. The relevant residues in the N-terminal region of Lcn-α are G1, T2, W3, D4, D5, Q8, G9, I10, R12, and V13 (Fig. (Fig.1A1A).
The lactococcin G immunity protein did not protect the two strains efficiently against enterocin 1071 (i.e., Ent1071A and Ent1071B) or the L. lactis MG1363 strain against the Ent1071A+Lcn-β combination (Table (Table3).3). This was expected, since both the Ent1071A+Ent1071B and the Ent1071A+Lcn-β combinations lack all of the above-mentioned residues that are present in the N-terminal region of Lcn-α. Unexpectedly, the lactococcin G immunity protein did, however, protect Lactococcus strain LMGT-2077 against the Ent1071A+Lcn-β combination (Table (Table3).3). Residues on the C-terminal side of residue 13 in Lcn-β (because the immunity protein protected against Ent1071A+Lcn-β but not against Ent1071A+β[1-13]/B[11-35] [Table [Table3])3]) apparently compensate for the lack of the above-mentioned residues in the N-terminal region of Lcn-α. Residues on the C-terminal side of residue 13 in Lcn-β that are not present at the corresponding positions in Ent1071B are E14, I16, K17, F19, G20, A23, I24, D30, and I35. The two latter residues are probably not important for recognition, since similar residues are present at corresponding positions in Ent1071B (N instead of D and V instead of I) and earlier mutagenesis results revealed that D30-to-N and I35-to-V mutations in Lcn-β do not affect the activity of lactococcin G (24).
According to the recently proposed structural model of lactococcin G (Fig. (Fig.2),2), the sequences covering residues 1 to 13 and 14 to 24 in Lcn-α and Lcn-β, respectively, will become adjacent to each other in the active lactococcin G complex. These sequences or parts of these sequences, as predicted by the structure, will form one protein-interacting entity in lactococcin G, and this may explain how residues in both sequences are (directly or indirectly) recognized by the lactococcin G immunity protein. The proposed structural model thus accommodates the present results showing that the lactococcin G immunity protein recognizes residues in sequences covering positions 1 to 13 and 14 to 24 in Lcn-α and Lcn-β, respectively.
The enterocin 1071 immunity protein protected Lactococcus strain LMGT-2077 against all tested peptide combinations and the L. lactis MG1363 strain against only some of the combinations (Table (Table3).3). The immunity protein protected the latter strain against the Ent1071A+Ent1071B and Lcn-α+Lcn-β wild-type combinations but not against the Ent1071A+Lcn-β and the Ent1071A+β[1-13]/B[11-35] hybrid combinations, even though the four combinations were highly and equally toxic toward the strain (Table (Table3).3). This finding indicates (i) that the peptide-induced process that leads to cell toxicity is by itself not necessarily sufficient to trigger the functioning of the immunity proteins and (ii) that the two complementary peptides are recognized as one physical entity (since both the Lcn-α+Lcn-β and Ent1071A+Ent1071B wild-type combinations were recognized by the immunity protein but the Ent1071A+Lcn-β hybrid combination was not, even though the three combinations are equally toxic). It is noteworthy that the enterocin 1071 immunity protein protected the L. lactis MG1363 strain against the α[1-16]/A[14-39]+Lcn-β and α[1-16]/A[14-39]+β[1-13]/B[11-35] hybrid combinations but not the similar Ent1071A+Lcn-β and Ent1071A+β[1-13]/B[11-35] hybrid combinations (Table (Table3).3). It thus appears that, like the lactococcin G immunity protein, the enterocin 1071 immunity protein requires residues present in the N-terminal region (residues 1 to 13) of Lcn-α but absent in Ent1071A for the recognition of lactococcin G as well as lactococcin-enterocin hybrids when the immunity protein is present in the L. lactis MG1363 strain. The fact that the Ent1071A+Lcn-β and Ent1071A+β[1-13]/B[11-35] combinations were recognized differently by the two indicator strains indicates that the functionality of the enterocin 1071 immunity protein also depends on a cellular component.
It has also been shown previously that the functionality of the immunity proteins for the pediocin-like (class IIa) bacteriocins depends on a cellular component (10, 15, 16). This component is most likely the membrane-embedded MptC and/or MptD subunit of the mannose-PTS, with which the pediocin-like bacteriocins interact, thereby triggering the cognate immunity proteins to bind to the bacteriocin-mannose-PTS complex (9). Interactions between the pediocin-like bacteriocins and the MptC and/or MptD subunit apparently alter the conformation of the mannose-PTS in a manner that results in membrane leakage, and this leakage is blocked by the binding of cognate immunity proteins to the bacteriocin-mannose-PTS complex. The two-peptide bacteriocins may possibly function in an analogous manner. By interacting with an integrated membrane (transport) protein that functions as the bacteriocin receptor, two-peptide bacteriocins may induce membrane leakage, which in turn is blocked by the binding of the cognate immunity proteins to the bacteriocin-receptor complex. The functionality of the immunity proteins would then be dependent on a cellular component—the bacteriocin receptor—and the immunity proteins would then specifically recognize the cognate bacteriocins indirectly through the conformational alteration the bacteriocins induce in the receptor. One may speculate further that the conformational alterations induced by the bacteriocins may vary to some extent depending on modifications/mutations in the bacteriocins and that this variation may influence how optimally the immunity proteins are able to recognize the bacteriocins. Interestingly, the enterocin 1071 immunity protein protected Lactococcus strain LMGT-2077 against all tested peptide combinations, and one may speculate that the cellular component present in this strain has more optimal interaction with the different hybrid peptide combinations than the components present in the other strains analyzed.
This work has been supported by EMBIO (steering board for research in molecular biology, biotechnology, and bioinformatics at the University of Oslo) and the Norwegian Research Council.
We thank Morten Skaugen at the mass spectrometry/proteomics core facility at the Norwegian University of Life Sciences (Ås, Norway) for mass spectrometry analysis of the purified peptide samples.
Published ahead of print on 28 December 2009.