The Inhibitory Spectrum of Thermophilin 9 from Streptococcus thermophilus LMD-9 Depends on the Production of Multiple Peptides and the Activity of BlpG_St, a Thiol-Disulfide Oxidase

Abstract

The blp_St cluster of Streptococcus thermophilus LMD-9 was recently shown to contain all the genetic information required for the production of bacteriocins active against other S. thermophilus strains. In this study, we further investigated the antimicrobial activity of S. thermophilus LMD-9 by testing the susceptibility of 31 bacterial species (87 strains). We showed that LMD-9 displays an inhibitory spectrum targeted toward related gram-positive bacteria, including pathogens such as Listeria monocytogenes. Using deletion mutants, we investigated the contribution of the three putative bacteriocin-encoding operons blpD_St-orf2, blpU_St-orf3, and blpE_St-blpF_St (bac_St operons) and of the blpG_St gene, which encodes a putative modification protein, to the inhibitory spectrum and immunity of strain LMD-9. Our results present evidence that the blp_St locus encodes a multipeptide bacteriocin system called thermophilin 9. Among the four class II bacteriocin-like peptides encoded within the bac_St operons, BlpD_St alone was sufficient to inhibit the growth of most thermophilin 9-sensitive species. The blpD_St gene forms an operon with its associated immunity gene(s), and this functional bacteriocin/immunity module could easily be transferred to Lactococcus lactis. The remaining three Bac_St peptides, BlpU_St, BlpE_St, and BlpF_St, confer poor antimicrobial activity but act as enhancers of the antagonistic activity of thermophilin 9 by an unknown mechanism. The blpG_St gene was also shown to be specifically required for the antilisteria activity of thermophilin 9, since its deletion abolished the sensitivities of most Listeria species. By complementation of the motility deficiency of Escherichia coli dsbA, we showed that blpG_St encodes a functional thiol-disulfide oxidase, suggesting an important role for disulfide bridges within thermophilin 9.

Bacteriocins are small (20- to 60-residue) ribosomally synthesized and extracellularly released peptides displaying antimicrobial activity against species that may or may not be closely related to the producer bacteria (24). Generally, bacteriocin producer strains secrete multiple antimicrobial peptides, each exhibiting a specific inhibitory spectrum, which provides the producer strain with a competitive advantage by eliminating concurrent strains (or species) sharing the same ecological niche. Well-known examples are the production by Streptococcus mutans of mutacins (five mutacins have been identified so far), which are active against streptococcal species present in the human oral cavity (19, 23), and the production by Lactobacillus sakei LTH673 of sakacin P, which has the greatest activity against Carnobacterium and Listeria species (39), and sakacin Q, mainly active against Lactobacillus coryniformis (31). The last decade has seen a growing interest in the bacteriocins of lactic acid bacteria (LAB), particularly those active against clostridia, listeria, and enteropathogen species, because of their potential use as natural biopreservatives to protect food products against bacterial contamination.

Streptococcus thermophilus is of major importance for the food industry, since it is widely used for the manufacture of dairy products (with an annual market of around $40 billion), and it is considered the second most important industrial dairy starter after Lactococcus lactis (22). In S. thermophilus LMD-9, the production of bacteriocin-like inhibitory substances active against other strains of S. thermophilus has recently been shown to depend on the chromosomally encoded class II blp_St locus (15). The class II bacteriocins of LAB include non-posttranslationally modified peptides (36) and are further subdivided into three main subcategories: IIa, the pediocin-like bacteriocins with strong antilisterial effects, which contain a conserved N-terminal YGNGVXC sequence (12), and IIb, bacteriocins whose activities depend on the complementary activities of two peptides (α and β peptides) (18). All other nonmodified bacteriocins are classified as class IIc (36). Class II bacteriocins of LAB permeabilize the target cell membrane by the formation of poration complexes, which leads to the release of essential molecules and the dissipation of the proton motive force (20).

As is the case for the production of many bacteriocins by LAB, the Blp_St-related antimicrobial activity of S. thermophilus is regulated at the transcriptional level by a cell-density mechanism (quorum sensing) (15, 25). The pheromone precursor BlpC_St is processed downstream of a double-glycine (two-Gly) motif and secreted through a specific transport apparatus consisting of the ABC transporter BlpA_St and the accessory protein BlpB_St (15). The processed forms of BlpC_St (D9C-30 [30 amino acids {aa}] and D9C-19 [19 aa]) activate a signaling cascade involving the BlpH_St/BlpR_St two-component system, which triggers the transcription of the bacteriocin and immunity structural genes (15). These genes are organized into three operons (blpD_St-orf2, blpU_St-orf3, and blpE_St-blpF_St), each comprising putative bacteriocin genes (collectively named bac_St genes) and orf genes (15). The four Bac_St peptides BlpD_St, BlpU_St, BlpE_St, and BlpF_St contain a two-Gly leader, which is likely cleaved off during secretion by the BlpAB_St transport system. However, their individual functionalities and their putative interactions as a multipeptide bacteriocin remained hypothetical (15). The Orf peptides and the predicted mature Bac_St peptides, except for the mature peptide BlpF_St* (the asterisk represents the putative mature part of the peptide, i.e., after cleavage downstream of the first two-Gly motif), contain stretches of highly hydrophobic residues that are predicted to form transmembrane segments (Table 1). The predicted mature peptides BlpD_St*, BlpU_St*, and BlpE_St* contain several Gly-Gly, Ala-Gly, and Gly-Ala repeats, which is typical of two-component class IIb bacteriocins (Table 1). Additionally, BlpE_St*, BlpU_St*, and BlpD_St* share 28, 28, and 29% identity (44, 36, and 51% similarity) with the β peptide of lactococcin M, the α peptide of thermophilin 13, and the β peptide of brochocin C, respectively (30, 33, 35).

TABLE 1.

Primary sequence analysis of the Bac_St and Orf peptides encoded in the blpD_St-blpF_St region

Gene product	Size (aa)	Sequencea	% Hydrophobic amino acid residuesb	No. of TM segmentsc	pI
BlpDSt*	35	LSCdegmlavGGlGAvGGpwGAvGGvlvGAalyCf	48.6	1	3.7
BlpUSt*	53	gCswGGfakqgvatgvgnglrlgiktrtwqGAvAGAAGGAivGGvgyGAtCww	41.5	2	9.9
BlpESt*	56	rvnwerwgmCGAsvavGAsegfsaaAGGtaffigpyaigtGAvGAaiGGGvallgC	48.2	1	6.1
BlpFSt*	43	avClmpnvnnkegdplkdgwvsppryrsgeaypmvylpvCaim	30.2	0	6.2
Orf1	66		36.4	1	8.8
Orf2	51		47.0	2	10.0
Orf3	57		43.9	2	9.4
Orf4	55		50.9	2	10.1
Orf5	56		50.0	2	8.9
Orf6	133		51.9	4	10.1
Orf7	130		56.1	4	9.7

^aOnly given for Bac_St peptides. The cysteine residues (boldface, underlined) and the GG, AG, and GA repeats (boldface) are indicated.

^bResidues taken into account are A, I, L, F, W, and V.

^cNumber of transmembrane segments, as predicted with TMHMM 2.0 (27).

The predicted processed forms of the four Bac_St peptides contain two cysteine residues at their N- and C-terminal ends. Recent studies have highlighted a strong correlation between the toxicity of class IIa bacteriocins and the number of disulfide bridges formed, which might play a role in their structural stability (11, 37). Intriguingly, the blpG_St-blpX_St operon, which is also regulated by the BlpC_St pheromone, encodes a putative modification protein (BlpG_St) with a thioredoxin fold domain containing the conserved thioredoxin catalytic motif CXXC. In Bacillus subtilis, the BdbB thiol-disulfide oxidase was shown to be required for the production of active sublancin 168, a bacteriocin that contains two disulfide bridges (9). Although the blpG_St-blpX_St operon of S. thermophilus does not seem to be required for the intraspecies antimicrobial activity of strain LMD-9 (15), its potential role in the inhibition of other species has not been investigated.

The aim of this study was to identify the genetic determinants involved in the Blp_St-related antimicrobial activity and immunity of S. thermophilus LMD-9 by genetic dissection of the blp_St locus and heterologous expression. We first determined the inhibitory spectrum of strain LMD-9 against a large set of indicator species. Appropriate blp_St mutants were then constructed, and their antimicrobial activities and immunity phenotypes were investigated.

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth conditions.

The bacterial strains and plasmids used in the present study, except for the indicator bacteria used to determine the antimicrobial spectrum of S. thermophilus LMD-9, are listed in Table 2. Escherichia coli was grown with shaking at 37°C in Luria-Bertani (LB) broth (38). E. coli motility was assayed on swarm plates containing 1% Bacto tryptone (wt/vol), 0.5% NaCl (wt/vol), and 0.3% agar (wt/vol), as previously described (3). Swarm plates were inoculated with 5 μl of exponentially growing cultures (optical density at 600 nm [OD₆₀₀], 1.0) of E. coli and incubated at 37°C for 20 h. S. thermophilus was grown anaerobically (BBL GasPak systems; Becton Dickinson, Franklin Lakes, NJ) in M17 broth (Difco Laboratories Inc., Detroit, MI) with 1% (wt/vol) glucose (M17G broth) at 42°C. L. lactis was grown in M17 broth with 0.5% (wt/vol) glucose at 29°C without shaking. When appropriate, erythromycin was added to the media at the following concentrations: 250 μg/ml for E. coli, 2.5 μg/ml for S. thermophilus, and 5 μg/ml for L. lactis. Solid agar plates were prepared by adding 2% (wt/vol) agar to the media.

TABLE 2.

Bacterial strains and plasmids used in this study

Strain, plasmid, or primer	Characteristic(s)a	Source or reference
Strains
S. thermophilus
LMD-9	Wild type	15
LF105 (Δ1)	LMD-9 Δ(blpDSt-blpFSt)	15
LF106 (Δ7)	LMD-9 Δ(blpGSt-blpXSt)	15
LF109 (Δ2)	LMD-9 Δ(blpUSt-blpFSt)	This study
LF110 (Δ3)	LMD-9 Δ(blpDSt-orf6)	This study
LF109.1 (Δ9)	LMD-9 Δ(blpUSt-blpFSt) ΔblpDSt	This study
LF109.2 (Δ10)	LMD-9 Δ(blpUSt-blpFSt) Δ(blpDSt-orf1)	This study
LF111 (Δ4)	LMD-9 Δ(blpDSt-orf2)	This study
LF111.1 (Δ4/Δ5)	LMD-9 Δ(blpDSt-orf2) Δ(blpESt-blpFSt)	This study
LF112 (Δ5)	LMD-9 Δ(blpESt-blpFSt)	This study
LF113 (Δ6)	LMD-9 Δ(blpUSt-orf6)	This study
LF114 (Δ8)	LMD-9 ΔblpGSt	This study
L. plantarum
NCIMB8826	Wild type	NCIMBb
L. lactis
NZ3900	MG1363 derivative	7
E. coli
TG1	supE hsdΔ5 thi Δ(lac-proAB) F′[traD36 proAB+lacIqlacZΔM15]	38
EC1000	Kmr RepA+	26
AH50	MC1000 phoR Δara714 leu+phoA68	21
AH55	AH50 dsbA::kan-1	21
Plasmids
pGIBG001	Emr; pMG36e in which an NcoI restriction site has been introduced to allow the cloning of genes in translational fusion with the P32 expression cassette	15
pGILF003	Emr; pGIBG001 with a 0.69-kb insert containing the blpGSt ORF of S. thermophilus LMD-9 in translational fusion with the P32 expression cassette	This study
pMG36e	Emr; E. coli-S. thermophilus shuttle vector; contains the P32 promoter	40
pMG36eT	Emr; pMG36e with a 0.36-kb insert containing the 3′ end of the ldhL ORF and the transcriptional terminator of ldhL from L. plantarum WCFS1	This study
pGILF004	Emr; pMG36eT with a 0.6-kb insert containing the blpDSt-orf2 operon of S. thermophilus LMD-9 in transcriptional fusion with the P32 promoter	This study
pGILF005	Emr; pMG36eT with a 0.6-kb insert containing the blpDSt-orf2 operon and a 3.55-kb insert containing the blpABSt ORFs of S. thermophilus LMD-9 in transcriptional fusion with the P32 promoter	This study
pGhost9	Emr Ts	28
pGILF109	Emr Ts; pGhost9 derivative for the deletion of blpUSt-blpFSt	This study
pGILF109.1	Emr Ts; pGhost9 derivative for the in-frame deletion of blpDSt in strain LF109	This study
pGILF109.2	Emr Ts; pGhost9 derivative for the deletion of blpDSt-orf1 in strain LF109	This study
pGILF110	Emr Ts; pGhost9 derivative for the deletion of blpDSt-orf6	This study
pGILF111	Emr Ts; pGhost9 derivative for the deletion of blpDSt-orf2	This study
pGILF112	Emr Ts; pGhost9 derivative for the deletion of blpESt-blpFSt	This study
pGILF113	Emr Ts; pGhost9 derivative for the deletion of blpUSt-orf6	This study
pGILF114	Emr Ts; pGhost9 derivative for the deletion of blpGSt	This study

^aKm^r, Em^r, and Ts indicate resistance to kanamycin, resistance to erythromycin, and temperature-sensitive mutations in the RepA replication protein, respectively.

^bNCIMB, The National Collections of Industrial and Marine Bacteria, Ltd., Aberdeen, Scotland.

DNA techniques and transformation.

General molecular biology techniques were performed according to the instructions given by Sambrook et al. (38). Plasmids derived from pMG36e (40) and pGhost9 (28) were constructed in E. coli strains TG1 (38) and EC1000 (26), respectively. Electrotransformation of E. coli, S. thermophilus, and L. lactis was performed as described by Dower et al. (10), Blomqvist et al. (4), and Leenhouts et al. (27), respectively. The transformed S. thermophilus cells were immediately resuspended in 1 ml of M17G and were incubated anaerobically for 6 h at 37°C (pMG36e derivatives) or 29°C (pGhost9 derivatives). Chromosomal DNAs of S. thermophilus and Lactobacillus plantarum were prepared as described previously (13). PCRs were performed with Pfu DNA polymerase (Promega, Madison, WI) in a GeneAmp PCR system 2400 (Applied Biosystems, Lennik, Belgium).

Construction of the blpG_St expression vector.

The entire open reading frame (ORF) of blpG_St was amplified by PCR with primers BlpG1 (5′-AAACCATGGTGAAAAAAAGGACATTAACGC-3′) and BlpG2 (5′-CGGGTACCTTCTCTTTCGTACTCTCTTCG-3′), yielding a 0.69-kb fragment that was then digested with NcoI and KpnI (the restriction sites are underlined in the primer sequences) and translationally fused with the P32 expression signals into the similarly digested pGIBG001 plasmid (15). The resulting plasmid was designated pGILF003.

Construction of the vectors for heterologous expression of BlpAB_St and blpD_St-orf2 in L. lactis.

First, the ldhL transcriptional terminator from L. plantarum NCIMB8826 was PCR amplified using primers LdhTER1 (5′-AAAGCATGCACATCATCAACTTGAAGGG-3′) and LdhTER2 (5′-ATACTGCCCCCAATCATAAGTCCACG-3′), digested with SphI (the restriction site is underlined in the LdhTER1 sequence) and KpnI (the restriction site present in the sequence of ldhL), and cloned into the similarly digested pMG36e vector (40), yielding pMG36eT. In a second step, a fragment encompassing the blpD_St ribosome binding site and the blpD_St-orf2 region was PCR amplified with primers pMGopD1 (5′-AAAGTCGACAAATTTTAGGAGGTAGTTGC-3′) and pMGopD2 (5′-AACCTGCAGGCTAATTCTTTCTATACTGCC-3′), digested with SalI and PstI (the restriction sites are underlined in the primer sequences), and transcriptionally fused with the P32 promoter into the similarly digested pMG36eT vector, yielding plasmid pGILF004. Finally, a PCR fragment encompassing the blpA_St ribosome binding site and the blpAB_St genes, obtained with primers BlpAB1 (5′-AACCTGCAGGATAATTTGTGATGAAAGGG-3′) and BlpAB2 (5′-AAAGCATGCTTAGCCATCAGTAATTCTCC-3′), was digested with SbfI and SphI (the restriction sites are underlined in the primer sequences) and cloned between the corresponding sites of plasmid pGILF004, downstream of the blpD_St-orf2 operon. In the resulting plasmid, pGILF005, the blpD_St-orf2 operon and the blpAB_St genes formed an artificial operon structure whose transcription was controlled by the P32 promoter.

Construction of deletion mutants in the blp_St locus.

The deletion vectors were constructed in the thermosensitive pGhost9 vector by successively cloning two fragments of approximately 1 kb, corresponding to the upstream and downstream regions of the target gene(s), respectively. Deletions were performed by a two-step homologous-recombination process, as previously described (29). Both recombination steps (plasmid integration and excision) were confirmed by PCR using primers located upstream and downstream of the recombination regions. Table S1 in the supplemental material gives an overview of the strategy used for the construction of the different deletion vectors and the corresponding S. thermophilus mutant strains.

Analysis of the antimicrobial activities and immunities of LMD-9 derivatives.

The synthetic mature form of BlpC_St, named D9C-30 (H₂N-SGWMDYINGFLKGFGGQRTLPTKDYNIPQA-COOH) (purity > 95%), was purchased from Sigma-Genosys (Sigma-Genosys Ltd., Haverhill, United Kingdom). Two alternative methods were used to assay antimicrobial activity, as previously described (15). For the spot-on-lawn method, overnight cultures of the producer strains were diluted 100-fold in fresh M17G broth and incubated anaerobically at 42°C. At an OD₆₀₀ of 0.1, the synthetic D9C-30 peptide was added at a final concentration of 400 ng/ml, and the cultures were then reincubated for 2 h (final OD₆₀₀, 1.6). Small volumes (5 μl) of the induced cultures were then spotted directly on a 6-ml soft M17G layer (0.8% agar) containing 10⁸ CFU of indicator strains. The plates were incubated anaerobically at 42°C overnight before analysis of the inhibition zones surrounding the producer cells. For the overlay method (multilayer method), overnight cultures of S. thermophilus or L. lactis producer strains were diluted 100-fold in fresh M17G broth and incubated anaerobically at 42°C (S. thermophilus) or at 29°C (L. lactis). At an OD₆₀₀ of 1.0, 100 μl of the cultures was diluted 10⁶-fold in 6 ml of prewarmed soft M17G medium (0.8% agar) and uniformly poured on a petri dish containing a supporting layer of 25 ml solid M17G medium (2% agar). A second 6-ml soft M17G layer containing 400 ng/ml D9C-30 was poured on the layer of producer cells. Following growth of the producer strain (10 h of incubation), a third 6-ml layer of soft medium containing 10⁸ CFU of the indicator strain was added. The plates were incubated for 10 h before analysis of the inhibition zones surrounding the producer colonies. The composition of the growth medium used for the layer of indicator cells, as well as the incubation conditions, was adapted according to the indicator species: strains of Streptococcus salivarius were grown anaerobically in M17G broth at 42°C, strains of Enterococcus faecalis were grown in M17 with 0.5% (wt/vol) glucose at 37°C, strains of Lactobacillus (29°C, aerobically) and Clostridium (37°C, anaerobically) were grown in MRS broth (Difco), strains of Bacillus (29°C, aerobically, for Bacillus fusiformis, Bacillus cereus, Bacillus licheniformis, Bacillus mycoides, and Bacillus thuringiensis; 42°C, anaerobically, for Bacillus coagulans) were grown in LB broth, and strains of Staphylococcus, Enterobacter, Micrococcus luteus, Salmonella, and Listeria were grown in LB broth in the presence of oxygen at 37°C. The strains of Staphylococcus, Enterobacter, Salmonella, Listeria, B. fusiformis, B. cereus, B. licheniformis, B. mycoides, B. thuringiensis, Micrococcus luteus, Enterobacter gergovia, Escherichia hermannii, Pseudomonas aeruginosa, and Salmonella enterica subsp. enterica serovar Bredeney were from the bacterial collection of the Unité de Microbiologie of the Université catholique de Louvain (Louvain-la-Neuve, Belgium).

RESULTS

Inhibitory spectrum of S. thermophilus LMD-9.

To date, Blp_St-associated antimicrobial activity of S. thermophilus LMD-9 has been reported to inhibit the growth of other S. thermophilus strains (15). Here, we further investigated the inhibitory spectrum of strain LMD-9 upon expression of the blp_St locus (400 ng/ml of the BlpC_St* pheromone, D9C-30); the susceptibilities of a range of bacterial strains were assayed using a multilayer protocol, as described previously (15). In total, 87 strains belonging to 31 bacterial species were tested as indicator strains. Among these, growth inhibition was observed for 33 strains, including all strains of S. thermophilus except LMG7952 and strains belonging to the closely related species S. salivarius, E. faecalis, and L. lactis (Table 3). Other species of LAB involved in food fermentations, such as L. plantarum, L. sakei, Lactobacillus curvatus, Lactobacillus fermentum, Lactobacillus casei, and Pediococcus pentosaceus were found to be resistant. Interestingly, Lactobacillus delbrueckii subsp. bulgaricus and Lactobacillus helveticus, used in cofermentations with S. thermophilus for the production of yogurt and cheese, respectively, were also not inhibited (Table 3). None of the gram-negative species tested (E. gergovia, E. coli, E. hermannii, P. aeruginosa, and S. enterica subsp. enterica serovar Bredeney) displayed sensitivity to the antimicrobial compounds produced by S. thermophilus LMD-9 (data not shown). Similarly, no antimicrobial activity was detected against the more distantly related gram-positive species Staphylococcus aureus, Staphylococcus hominis, Clostridium acetobutylicum, B. cereus, B. thuringiensis, B. licheniformis, B. mycoides, and M. luteus (data not shown). Notably, a number of exceptions were observed: S. thermophilus LMD-9 was found to inhibit the growth of the gram-positive species B. coagulans, B. fusiformis, and a number of Listeria species, including all strains of Listeria innocua and Listeria seeligeri, and of the pathogenic species Listeria monocytogenes and Listeria ivanovii (Table 3).

TABLE 3.

Inhibitory spectrum of S. thermophilus LMD-9 and derivatives

Indicator species or strain	No. of strains tested	No. of indicator strains inhibited by the induced producer strainsa
		LMD-9 (D U E F) +	Δ1 (−) +	Δ2 (D) +	Δ3 (E F) +	Δ4/Δ5 (U) +	Δ4 (U E F) +	Δ9 (−) +	Δ5 (D U) +	Δ6 (D E F) +	Δ7 (D U E F) −b	Δ8 (D U E F) −b
Streptococcus
S. thermophilus	8	7	0	7	0	0	0	0	7	7	7	7
S. thermophilus SFi16	1	1	0	0	0	0	0	0	0	0	1	1
S. thermophilus LMG18311	1	1	0	1	1d	0	1d	0	1	1	1	1
S. salivarius	3	3	0	3	0	0	0	0	3	3	3	3
Enterococcus
E. faecalis OG1X	1	1	0	1	0	0	0	0	1	1	1	1
E. faecalis JH2-SS	1	1	0	1d	0	0	0	0	1d	1	1d	1d
E. faecalis JH2-2	1	1	0	1	0	0	0	0	1	1	1d	1d
Lactococcus
L. lactis	3	3	0	3	0	0	0	0	3	3	3	3
L. lactis IL-1403	1	1	0	0	0	0	0	0	1	0	1	1
L. lactis MG1363	1	1	0	1d	0	0	0	0	1d	1d	1d	1d
Bacillus
B. coagulans	4	3	0	3	0	0	0	0	3	3	3	3
B. coagulans NCIB8523	1	1	0	1	0	0	0	0	1	1	1d	1d
B. fusiformis SI0056	1	1	0	1	0	0	0	0	1	1	1	1
Listeria
L. monocytogenes	3	3	0	3	0	0	0	0	3	3	0	0
L. monocytogenes IHE-92	1	1	0	1	0	0	0	0	1	1	1d	1d
L. ivanovii IHE-3	1	1	0	1	0	0	0	0	1	1	1	1
L. innocua IHE-2	1	1	0	1	0	0	0	0	1	1	0	0
L. seeligeri	2	2	0	2	0	0	0	0	2	2	0	0
L. welshimeri	3	0	NTc	NT	NT	NT	NT	NT	NT	NT	NT	NT
Lactobacillus
L. delbrueckii subsp. bulgaricus	2	0	NT	NT	NT	NT	NT	NT	NT	NT	NT	NT
L. helveticus	3	0	NT	NT	NT	NT	NT	NT	NT	NT	NT	NT

^aAntimicrobial activity was assayed using the overlay protocol in the presence of 400 ng/ml D9C-30. Growth of the indicator strain was considered inhibited when the radius of the inhibitory zone was at least 1.5 mm. Each experiment was repeated at least three times. The letters in parentheses indicate the content in bac_St genes of the producer strain: D, blpD_St; U, blpU_St; E, blpE_St; F, blpF_St. “−” denotes the absence of bac_St genes. The presence (+) or absence (−) of blpG_St, encoding the putative disulfide oxidase, is also indicated.

^bIn strain Δ7, the blpG_St-blpX_St operon is deleted, while blpG_St alone is deleted in strain Δ8.

^cNT, not tested.

^dHighly significant difference in the size of the inhibitory zone produced by the LMD-9 derivative compared to the wild-type LMD-9 strain (Student's t test with a P value of <0.01).

Thermophilin 9 is a multipeptide bacteriocin.

The blpD_St-blpF_St region is organized into three operons encoding a total of four Bac_St peptides (BlpD_St, BlpU_St, BlpE_St, and BlpF_St) that show characteristics of class IIb bacteriocins (15). To confirm the implications of the three operons in the antimicrobial activity of strain LMD-9, the inhibitory spectrum of the Δ(blpD_St-blpF_St) mutant (Δ1) (Fig. 1A and Table 3) was investigated. As expected, none of the 33 LMD-9-sensitive strains was inhibited. This shows that the region encompassing the three operons is responsible not only for the intraspecies inhibition, as previously shown for S. thermophilus LMG18311 and CNRZ1066 (15), but also for the interspecies antimicrobial activity of LMD-9. The relative contributions of the different bac_St genes to this antimicrobial activity was then investigated by analyzing the inhibitory spectra of three mutant strains (Δ2, blpD_St⁺; Δ3, blpE_St⁺ blpF_St⁺; Δ4/Δ5, blpU_St⁺) (Fig. 1A), each carrying only one of the three bac_St operons, under conditions of induction of the blp_St locus (400 ng/ml of D9C-30). This assay was performed exclusively with the 33 indicator strains that showed sensitivity to the wild-type LMD-9 strain (Table 3).

FIG. 1.

Schematic representation of the genetic deletions performed in this study. Complete strain names and the corresponding genotypes are listed in Table 2. Genes encoding peptides with known or predicted functions are represented by patterned arrows: ABC transporter (small squares), accessory transport protein (large squares), induction factor (dark gray), response regulator (vertical lines), histidine kinase (horizontal lines), hydrophobic peptides of unknown function (black), hydrophilic peptides of unknown function (white), peptide similar to immunity proteins of class IIa bacteriocins (checkerboard squares), and modification protein (dots). Genes encoding the putative bacteriocins with a two-Gly leader are represented by light-gray arrows. The flags represent promoters, black flags for BlpC_St-induced promoters and gray flags for vegetative promoters. The hairpin structures indicate potential transcription terminators. The letters and numbers in italics refer to the corresponding blp_St and orf genes, respectively. (A) Deletions performed in the blp_St locus of strain LMD-9. (B) Deletions performed in the blp_St locus of strain LF109 [LMD-9 Δ(blpU_St-blpF_St)].

Strain Δ2, which contains only the blpD_St-orf2 operon, retained most of its antimicrobial activity; the cell target range and activity were only slightly attenuated compared to those of the wild-type strain (31/33 strains were inhibited). Growth inhibition of S. thermophilus SFi16 and L. lactis IL-1403 was lost, while the sizes of the inhibitory zones for the indicator strains L. lactis MG1363 and E. faecalis JH2-SS were significantly reduced (Tables 3 and 4). In contrast, the sole expression of the blpU_St-orf3 operon (strain Δ4/Δ5) failed to inhibit the growth of any of the 33 indicator strains (Table 3). The blpE_St-blpF_St operon alone (strain Δ3) conferred a poor antimicrobial activity against S. thermophilus LMG18311, while growth of the remaining 32 indicator strains was not affected (Tables 3 and 4). These results strongly suggest that blpD_St encodes the key determinant of the antagonistic activity of strain LMD-9. To further validate this hypothesis, we determined the inhibitory spectra of strains Δ(blpD_St-orf2) (Δ4, blpU_St⁺ blpE_St⁺ blpF_St⁺) and ΔblpD_St Δ(blpU_St-blpF_St) (Δ9, bac_St mutant and orf1⁺ orf2⁺) (Fig. 1A and B). As expected, neither of these two mutant strains inhibited growth of any of the LMD-9-sensitive strains, with the exception of S. thermophilus LMG18311, which displayed sensitivity to strain Δ4 (Table 3 and 4).

TABLE 4.

Antimicrobial activities of S. thermophilus LMD-9 and derivatives against a defined set of indicator strains

Indicator strain	Inhibitory zone produced by the induced producer strainsa
	LMD-9 (D U E F) +	Δ2 (D) +	Δ3 (E F) +	Δ4 (U E F) +	Δ5 (D U) +	Δ6 (D E F) +	Δ7 (D U E F) −	Δ8 (D U E F) −
S. thermophilus LMG18311	6.5 ± 0.5	6.5 ± 0.5	3.5 ± 0.5c	3.6 ± 0.4c	6.7 ± 0.4	6.5 ± 0.5	6.5 ± 0.5	6.5 ± 0.5
E. faecalis JH2-2	3.5 ± 0.4	3.4 ± 0.3	NDb	ND	3.4 ± 0.4	3.5 ± 0.4	1.8 ± 0.2c	1.9 ± 0.2c
E. faecalis JH2-SS	3.5 ± 0.5	2.0 ± 0.1c	ND	ND	2.0 ± 0.1c	3.5 ± 0.5	1.9 ± 0.2c	1.8 ± 0.2c
L. lactis subsp. cremoris MG1363	3.0 ± 0.2	2.0 ± 0.1c	ND	ND	2.0 ± 0.1c	2.0 ± 0.1c	1.9 ± 0.2c	1.9 ± 0.2c
B. coagulans NCIMB8523	3.0 ± 0.1	3.0 ± 0.2	ND	ND	3.0 ± 0.1	3.1 ± 0.2	2.0 ± 0.2c	2.0 ± 0.1c
L. monocytogenes IHE-92	3.3 ± 0.4	3.3 ± 0.3	ND	ND	3.4 ± 0.4	3.3 ± 0.4	2.4 ± 0.3c	2.2 ± 0.2c

^aAntimicrobial activity was assayed using the overlay protocol, in the presence of 400 ng/ml D9C-30. The numbers represent the radii of the inhibitory zones around the producer colonies measured around at least 60 colonies (20 colonies from three independent experiments). The results are given as mean values ± standard deviations. The letters in parentheses indicate the content in bac_St genes of the producer strain: D, blpD_St; U, blpU_St; E, blpE_St; F, blpF_St. The presence (+) or absence (−) of blpG_St, encoding the putative disulfide oxidase, is also indicated.

^bND, not detected.

^cHighly significant difference in the size of the inhibitory zone produced by the LMD-9 derivative compared to the wild-type LMD-9 strain (Student's t test with a P value of <0.01) (Table 3).

The attenuated inhibitory spectrum and activity of strain Δ2 (blpD_St⁺) suggest that the growth inhibition of four LMD-9-sensitive strains (S. thermophilus SFi16, L. lactis IL-1403, L. lactis MG1363, and E. faecalis JH2-SS) relies on the coexpression of blpD_St together with blpU_St and/or blpEF_St. Strains Δ(blpE_St-blpF_St) (Δ5, blpD_St⁺ blpU_St⁺) and Δ(blpU_St-orf6) (Δ6, blpD_St⁺ blpE_St⁺ blpF_St⁺) were thus constructed in order to compare their inhibitory spectra to those of strains LMD-9 and Δ2 (blpD_St⁺). The results showed that the growth inhibition of L. lactis IL-1403 could occur only when both the blpD_St-orf2 and blpU_St-orf6 operons were expressed (Δ5) (Table 3), while the inhibition of E. faecalis JH2-SS required the coexpression of the blpD_St-orf2 and blpE_St-blpF_St operons (Δ6) (Tables 3 and 4). Interestingly, the complete inhibition of S. thermophilus SFi16 and L. lactis MG1363 required the expression of all three bac_St operons (Tables 3 and 4).

Taken collectively, our results demonstrate that blpD_St is the main determinant of antimicrobial activity in S. thermophilus LMD-9 and that the coexpression of blpU_St and blpEF_St with blpD_St enhances the activity and cell target range of BlpD_St*. The Bac_St peptides of S. thermophilus LMD-9 thus seem to be part of a multipeptide bacteriocin, which was named thermophilin 9.

BlpG_St is a functional disulfide oxidase required for the antilisterial activity of thermophilin 9.

We have shown previously that the blpG_St-blpX_St operon is not associated with intraspecies inhibition (15). Here, we investigated a possible role of these genes in the inhibition of indicator strains belonging to other species. Deletion of this operon [strain Δ(blpG_St-blpX_St), Δ7] (Fig. 1A) completely abolished the growth inhibition of six of the eight LMD-9-sensitive Listeria strains (Table 3). In addition, this mutant strain displayed decreased antimicrobial activity against five indicator strains (E. faecalis JH2-SS and JH2-2, L. monocytogenes IHE-92, B. coagulans NCIB8523, and L. lactis MG1363), as measured by the size of the inhibitory zone (Table 4). The role of the putative modification protein BlpG_St in the attenuated interspecies inhibitory activity of the mutant strain Δ7 was further investigated by constructing an in-frame deletion of blpG_St alone (Δ8) (Fig. 1A). As shown in Tables 3 and 4, the two mutant strains exhibited identical inhibitory spectra, thus revealing the specific role of BlpG_St in interspecies inhibition.

Besides the CYYC thioredoxin motif (Fig. 2A), BlpG_St (230 aa) contains a putative 27-aa signal sequence with a predicted cleavage probability of 1.0 (prediction by signalP). This suggests that the protein is released at the cell surface, where it might promote the formation of disulfide bonds in the antimicrobial compounds secreted by S. thermophilus LMD-9. To validate this hypothesis, we tested the ability of blpG_St to complement the motility deficiency of a dsbA-deficient mutant of E. coli AH50 (21), as previously described (3). DsbA is a periplasmic disulfide oxidase that catalyzes the formation of disulfide bridges in proteins of the flagellar machinery of E. coli (6). A blpG_St expression vector was constructed by cloning the blpG_St ORF downstream of the P32 expression cassette carried by an autoreplicative plasmid. As shown in Fig. 2B, expression of blpG_St was able to restore motility in the dsbA-deficient background, showing that BlpG_St is exported in the periplasmic space and is a functional thiol-disulfide oxidase.

FIG. 2.

Functional role of BlpG_St. (A) Primary sequence alignment of the putative thioredoxin domains of BlpG_St from S. thermophilus LMD-9 and similar proteins (accession numbers: S. mutans UA159 Smu.1904c, NP722210; Brochothrix campestris ATCC 43754 BrcD, AAC95141; Lactobacillus salivarius UCC118 Orf1, AAM61773; B. subtilis 168 BdbB, CAB14062; E. coli K-12 DsbA, AAB02995). Blocks of conserved identical residues are shown on a black background. Amino acids showing conservation among all proteins are shown on a gray background. (B) Complementation of the motility deficiency of E. coli AH50 dsbA by expression of BlpG_St. Motility assays were performed on 0.3% agar-containing plates. The plates were incubated at 37°C for 20 h.

The functional role of orf genes in the innate immunity of S. thermophilus LMD-9.

In a previous study, we demonstrated that the genetic determinants of immunity against thermophilin 9 are located within the blpD_St-blpF_St region (15). Each bac_St gene was shown to be cotranscribed with one or more orf genes (Fig. 1A). These orf genes encode small peptides that display structural similarities (high pI and the presence of one to four predicted transmembrane segments) (Table 1) to the immunity peptides of the class IIb bacteriocins brochocin C and lacticin F (2, 33), supporting their potential roles in thermophilin 9 immunity. The absence of a signal peptide and a two-Gly leader sequence suggests that the products of the orf genes are located in the cell membrane. To investigate which of these orf genes is required for immunity, we first determined the impact of deleting each bac_St operon individually: the immunity phenotypes of mutants Δ4 (orf3⁺ to orf7⁺), Δ5 (orf1⁺ to orf6⁺), and Δ6 (orf1⁺ orf2⁺ orf7⁺) were determined against the induced thermophilin 9-producing strain LMD-9 using the spot-on-lawn method. Among the three indicator strains tested, only strain Δ6 displayed resistance against LMD-9 (Table 5), suggesting that orf7, together with orf1 and/or orf2, encodes the immunity determinants of thermophilin 9.

TABLE 5.

Immunities of S. thermophilus LMD-9 and derivatives

Indicator strain	No. of orf genesb	Producer straina
		LMD-9 (D U E F)	Δ4 (U E F)	Δ6 (D E F)	Δ5 (D U)	Δ2 (D)
LMD-9c	1-7	+	+	+	+	+
Δ1	−	−	+	−	−	−
Δ6	1, 2, 7	+	+	+	+	+
Δ4	3-7	−	+	−	−	−
Δ5	1-6	−	+	+	+	+
Δ3	7	−	+	−	−	−
Δ2	1, 2	−	+	+	+	+
Δ9	1, 2	−	+	+	+	+
Δ10	2	−	+	−	−	−

^aThe immunity phenotype of the strains was determined by the spot-on-lawn method. “+” denotes the ability of the indicator strain to grow in the presence of the producer strain. The letters in parentheses indicate the content in bac_St genes of the producer strain: D, blpD_St; U, blpU_St; E, blpE_St; F, blpF_St.

^bContent in orf genes of the producer strain. “−” denotes the absence of orf genes.

We further investigated the immunity specificities of the orf1, orf2, and orf7 genes by comparing the resistances of strains Δ1 (no orf gene), Δ2 (orf1⁺ orf2⁺), and Δ3 (orf7⁺) against a set of LMD-9 derivatives producing only a subset of Bac_St peptides: Δ4 (blpU_St⁺ blpE_St⁺ blpF_St⁺), Δ6 (blpD_St⁺ blpE_St⁺ blpF_St⁺), Δ5 (blpD_St⁺ blpU_St⁺), and Δ2 (blpD_St⁺). As shown in Table 5, all strains were able to grow when tested against strain Δ4, with the blpD_St-orf2 operon deleted, which supports the previous observation that BlpU_St* and BlpE_St* BlpF_St* have no or poor antimicrobial activity, respectively. Reciprocally, expression of the blpD_St-orf2 operon alone (indicator strain Δ2) was sufficient to confer immunity to all producer strains tested (Δ2, Δ4, Δ5, and Δ6), provided that all four bac_St genes were not expressed together (producer strain LMD-9; Table 5). Since BlpD_St* is the main component of thermophilin 9 activity, these results suggest that orf1 and/or orf2 is specifically required for immunity against BlpD_St*, while the concomitant expression of orf7 is required against the antimicrobial activity resulting from a combination of BlpD_St* with BlpU_St* and BlpE_St* BlpF_St*.

To specifically assess the function of orf1 in immunity to BlpD_St*, the resistance spectra of strains bearing either both orf1 and orf2 genes [ΔblpD_St Δ(blpU_St-blpF_St), Δ9] (Fig. 1B) or orf2 alone [Δ(blpD_St-orf1) Δ(blpU_St-blpF_St), Δ10] (Fig. 1B) were compared. As expected, strain Δ9 displayed a resistance spectrum identical to that of strain Δ2 (Table 5), showing that the BlpD_St* immunity determinants were efficiently expressed in a background devoid of blpD_St. The absence of orf1 (strain Δ10) resulted in the complete loss of immunity against all strains producing BlpD_St* (Table 5), confirming that at least orf1 is required in this process.

Heterologous production of Blp_St bacteriocins.

The main determinants of the antimicrobial activity and immunity of thermophilin 9 were shown to be encoded within the blpD_St-orf2 operon. In order to confirm these results, we tested the possibility of conferring BlpD_St-dependent antimicrobial activity to L. lactis by constitutive expression of the blpD_St-orf2 operon. The operon was cloned in transcriptional fusion with the vegetative promoter P32 carried by the multicopy plasmid pMG36eT (40). Introduction of this construct (plasmid pGILF004) into L. lactis NZ3900 did not result in the production of antimicrobial compounds active against the indicator strain S. thermophilus LMG18311, as assayed with the overlay protocol (Fig. 3). Growth inhibition of S. thermophilus LMG18311 could be achieved only through the coexpression of the blpD_St-orf2 operon with the transporter genes blpAB_St in an artificial operon structure (plasmid pGILF005). L. lactis NZ3900(pGILF005) was active against the same strain carrying the cloning vector pMG36eT and against all LMD-9-sensitive S. thermophilus strains (data are shown only for LMG18311) (Fig. 3), except SFi16. This is consistent with the previous observation that all three bac_St operons are required for inhibition of strain SFi16 (Table 3). As for immunity, plasmid pGILF004 conferred resistance to the LMD-9 derivatives Δ5 (blpD_St⁺ blpU_St⁺) and Δ6 (blpD_St⁺ blpE_St⁺ blpF_St⁺), but not against LMD-9 itself (blpD_St⁺ blpU_St⁺ blpE_St⁺ blpF_St⁺) (data not shown). The immunity spectrum of NZ3900(pGILF005) was thus identical to that of the LMD-9 mutant bearing the blpD_St-orf2 operon alone (Δ2), confirming the finding that orf1 and orf2 are sufficient to protect against BlpD_St*.

FIG. 3.

Antimicrobial activity of L. lactis NZ3900 containing the blpD_St-orf2 (pGILF004) or the blpD_St-orf2 and blpAB_St expression vector (pGIL005) against the indicator strain S. thermophilus LMG18311 (overlay method).

These results demonstrate that the blpD_St-orf2 operon is a bacteriocin/immunity-encoding module able to fulfill its functions in the absence of other genes from S. thermophilus. However, efficient secretion and maturation of BlpD_St require a specific transport system encoded by blpAB_St, which seems to have no functional counterpart in L. lactis.

DISCUSSION

S. thermophilus LMD-9 displays an inhibitory spectrum against gram-positive bacteria (10 sensitive species among 26 gram-positive species tested) and is mainly active against closely related species. Its cell target range was shown to depend on the secretion of at least three Bac_St peptides and the activity of BlpG_St, a thiol-disulfide oxidase specifically required for the inhibition of Listeria species.

Among the bac_St genes, the sole expression of blpD_St was sufficient to inhibit most indicator strains. Its coexpression with the blpU_St- and blpE_St-blpF_St-containing operons further increased the antimicrobial activity and cell target range of S. thermophilus LMD-9. The results obtained from our deletion analysis suggest that thermophilin 9 is a multicomponent bacteriocin whose activity results from the combination of BlpD_St* with BlpU_St*, BlpE_St*, and/or BlpF_St*. In this context, the BlpD_St* and BlpU_St* BlpE_St* BlpF_St* peptides, respectively, could be compared to the active (α) and nonactive (β) subunits of thermophilin 13 (30), brochocin C (33), or lactacin F (1), all belonging to the class IIb bacteriocins. Interestingly, coexpression of the genes encoding the α and β peptides of lacticin F from Lactobacillus johnsonii was also shown to expand its cell target range (1). For thermophilin 9, the combination of Bac_St peptides required for optimal growth inhibition appears to depend on the indicator species and strain. The efficient antimicrobial activity of thermophilin 9 could result from the formation of pores consisting of the α peptide BlpD_St* and one or more of the β peptides BlpU_St*, BlpE_St*, or BlpF_St*, as proposed for most two-component bacteriocins (18). Alternatively, the different Bac_St peptides could act synergistically to form pores of complementary specificities, resulting in increased efficiency, similar to what has been hypothesized for the one-component lactococcins A and B (class IIc) and the two-component lactococcin M (class IIb) of L. lactis (35).

To our knowledge, BlpG_St from S. thermophilus LMD-9 and BdbB from B. subtilis 168 (9) are the only examples of disulfide oxidases whose involvement in bacteriocin production has been demonstrated. However, their direct implication in the formation of disulfide bridges in the associated bacteriocin has never been shown. With the exception of the conserved thioredoxin motif (CXXC), the catalytic domain of BlpG_St differs significantly from those of BdbB (15% identity) and DsbA (30% identity). Higher similarities were observed to the catalytic domains of the putative thioredoxins BrcD (45% identity), Smu.1904c (64% identity), and Orf1 (60% identity), encoded in the class IIb bacteriocin loci involved in the production of brochocin C (33), mutacin V (19), and ABP-118 (14), respectively (Fig. 2A). Interestingly, these bacteriocins share structural features with BlpD_St*, BlpU_St*, and BlpE_St*: these GA-rich peptides are predicted to form amphiphatic/hydrophobic α helices and to contain N- and C-terminal cysteine residues. Recently, the presence of one disulfide bridge was reported for BrcA (an α peptide of brochocin C) (17). In the present study, a correlation was observed between the presence of blpG_St in the producer strain and the antilisterial activity of thermophilin 9 (Table 3). Since only BlpD_St* seems to be required for the inhibition of Listeria strains, it can be speculated that BlpG_St is required for the formation of intra- and/or intermolecular disulfide bridges between the cysteine residues of BlpD_St*. These disulfide bridges might be important for the interaction of BlpD_St* with the membrane in Listeria species, whose lipid composition is known to differ significantly from that of LAB membranes (16).

Different mechanisms of immunity against bacteriocins have been reported. The immunity factors can either interact with the bacteriocin itself (e.g., NisI), thus preventing its interaction with the membrane, or actively export the intracellularly accumulated bacteriocins (e.g., NisFEG or Smu1913 in S. mutans) (5, 32). Recently, a novel mechanism was reported for class IIa and class IIc bacteriocins, in which the immunity peptides bind to the complex involving the bacteriocin and its membrane receptor, the mannose phosphotransferase (PTS) system (8). However, this PTS system does not seem to be involved in binding of thermophilin 9 to the target membrane, since a mutant of the indicator strain S. thermophilus LMG18311 in which the mannose PTS system was inactivated displayed the same level of sensitivity to the LMD-9 producer strain (data not shown). The structural similarities between Orf7 from S. thermophilus LMD-9 and Smu1913 from S. mutans (four predicted transmembrane domains and a large number of basic residues at the C-terminal end) might suggest a similar mode of action. Immunity against BlpD_St* could also occur through direct interactions of the Orf1, Orf2, and/or Orf7 peptide with BlpD_St*, which displays opposite charge properties (Table 1).

In conclusion, our results show that the blpD_St-orf2 operon encodes a functional bacteriocin/immunity module. A similar functional organization could be hypothesized for the blpU_St-orf3 and blpE_St-blpF_St operons, although the cell target range of BlpU_St* and BlpE_St*/BlpF_St* remains to be identified. The S. thermophilus strains used as indicator strains in this study all carry blp_St loci, which contain the blpABC_St and blpRH_St operons, as well as blpG_St (data not shown). The region located between blpR_St and blpG_St, which contains the bacteriocin/immunity determinants in strain LMD-9, is highly variable in size, between 2.5 kb (SFi16) and 4.5 kb (LMG7952). These strains also differ in the number and nature of the bacteriocin/immunity modules present in their blp_St loci: most strains carry only one bacteriocin operon (similar to blpD_St-orf2 or blpU_St-orf3), while LMG7952 bears two of these modules (similar to blpD_St-orf2 and blpE_St-blpF_St). The presence of highly similar sequences upstream (blpRH_St) and downstream (blpG_St) of the bacteriocin/immunity modules, as well as the high sequence conservation between the promoter regions of all bac_St operons, suggests a plasticity mechanism for blp_St loci by the acquisition/loss of modules through homologous recombination, as proposed for the sakacin P cluster of L. sakei (34). The insertion element present in all sequenced clusters except that of LMD-9 might also play a role in this process. Finally, the modular nature of the blp_St locus offers interesting perspectives for the engineering of industrial LAB with the specific aim of inhibiting the growth of nonstarter strains or food-borne pathogen species, such as listeria.