SMU.152 Acts as an Immunity Protein for Mutacin IV

Abstract

Streptococcus mutans, a principal causative agent of dental caries, secretes antimicrobial peptides known as mutacins to suppress the growth of competing species to establish a successful colonization. S. mutans UA159, a sequenced strain, produces at least two major mutacins, mutacins IV and V. Mutacin IV is a two-peptide mutacin encoded by nlmAB genes, which are mapped just upstream of a putative immunity-encoding gene SMU.152. Here we explored the function of SMU.152 as an immunity protein. We observed that overexpression of SMU.152 in two sensitive host strains converted the strains to become immune to mutacin IV. To identify the residues that are important for immunity function, we sequentially deleted residues from the C-terminal region of SMU.152. We observed that deletion of as few as seven amino acids, all of which are highly charged (KRRSKNK), drastically reduced the immunity function of the protein. Furthermore, we identified two other putative immunity proteins, SMU.1909 and SMU.925, which lack the last four charged residues (SKNK) that are present in SMU.152 but contain the KRR residues. Synthetic addition of SKNK residues to either SMU.1909 or SMU.925 to reconstitute the KRRSKNK motif and expressing these constructs in sensitive cells rendered the cells resistant to mutacin IV. We also demonstrated that deletion of Man-PTS system from a sensitive strain made the cells partially resistant to mutacin IV, indicating that the Man-PTS system plays a role in mutacin IV recognition.

INTRODUCTION

Bacteriocins are ribosomally synthesized small antimicrobial peptides largely produced by lactic acid bacteria and are active against a wide range of bacterial species (8, 27). Bacteriocins are generally divided into two main categories (10, 30): the class I bacteriocins, commonly known as lantibiotics, which contain posttranslationally modified peptides with circular structure forming lanthionine and β-methyllanthionine residues, and the class II nonmodified peptide bacteriocins. Bacteriocins belonging to class II are additionally divided into four subgroups (IIa to IId) (10). Pediocin-like linear peptide bacteriocins with very similar amino acid sequences are grouped under class IIa, two-peptide linear bacteriocins are placed under class IIb, the cyclic unmodified bacteriocins belong to class IIc, and the nonpediocin one-peptide linear bacteriocins are classified under class IId (10). The unmodified two-peptide bacteriocins belonging to class IIb contain two different peptides, and both peptides must be present in equimolar amounts to exert optimal antimicrobial activity (31, 32). After the first isolation of a two-peptide bacteriocin, lactococcin G (29), several two-peptide bacteriocins were isolated and characterized (31). The genes encoding the two-peptide bacteriocins are always found next to each other, along with a gene that encodes a membrane associated immunity protein that seems to protect the bacteriocin producer strain from being killed by its own bacteriocin.

Mechanisms by which immunity proteins confer protection to the producer strains can vary widely. For example, immunity against nisin, the best-characterized lantibiotics (class I), involves a membrane-associated protein, NisI, which sequesters the bacteriocins, and an ABC transporter complex (NisFEG) that expels the bacteriocin from the cell (37, 38). For some pediocin-like (class IIa) and some class IIc bacteriocins, the proteins that confer immunity have been shown to directly bind to the bacteriocin receptor and therefore block the pore formation by the bacteriocin (12). Recently, it has been shown that Abi family proteins consist of membrane-bound putative metalloproteases could confer protection against two-peptide bacteriocins such as plantaricins (PlnEF and PlnJK) produced by Lactobacillus plantarum (23). Furthermore, Abi family proteins display extensive cross-immunity against noncognate two-peptide bacteriocins. Because Abi encodes putative metalloproteases, it has been proposed that Abi-mediated immunity could involve direct proteolytic degradation of the bacteriocins (23).

Streptococcus mutans, a lactic acid bacterium, is one of the primary causative agents of dental caries (18, 24). Occasionally, S. mutans can enter the bloodstream and cause transient bacteremia and infective endocarditis (24). The organism possesses several properties allowing it to successfully colonize and survive in the human host. In the oral cavity, the bacterium can form dental plaque, a type of biofilm, on the tooth surfaces (5). Dental plaque is home to one of the most complex bacterial floras associated with human body. Thus far, more than 700 different bacterial species have been identified from human oral cavity, and the majority of them are associated with dental plaque (1, 11). However, S. mutans has ability to outcompete other bacteria by producing various bacteriocins known as mutacins (3, 4). Based on the bactericidal activity, most of the mutacins are also classified into five major types, I through V (28). The sequenced S. mutans strain UA159 produces at least two mutacins (IV and V) and encodes six other small peptides with high degree of similarity with bacteriocins (2, 17). While mutacin IV displays antimicrobial spectrum against various streptococcal species (19, 36), mutacin V appears to have a wide spectrum antimicrobial activity, ranging from mitis streptococci to lactococci and micrococcus (17).

Mutacin IV is a class IIb bacteriocin and is encoded by the nlmA and nlmB genes, which are organized in an operon. Both NlmA (44-residue) and NlmB (49-residue) peptides contain a single GxxxG motif, which is essential for the bacteriocin activity due to its involvement in helix-helix interaction (32). Both peptides also contain signal sequences with a well-conserved GG motif, the site where peptidase cleavage occurs during export. NlmTE, a transporter complex, cleaves the signal sequence and mediates the export of these peptides (16).

Immediately downstream of the nlmAB operon lies another gene, SMU.152, which is annotated in the GenBank database as an immunity protein. In the present study, we characterized this gene and showed that SMU.152 can confer protection against mutacin-IV in Streptococcus gordonii, an indicator strain traditionally used for mutacin IV activity. Furthermore, we show that the last seven charged amino acid residues at the C-terminal domain of the protein are essential for optimum immunity.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Escherichia coli strain DH5α was grown in Luria-Bertani medium supplemented, when necessary, with ampicillin (Ap; 100 μg/ml), erythromycin (Em; 300 μg/ml), and kanamycin (Km; 100 μg/ml). S. mutans UA159 and S. gordonii DL-1 were grown in Todd-Hewitt medium (BBL; Becton Dickson) supplemented with 0.2% yeast extract (THY) at 37°C. When necessary, erythromycin (5 to 10 μg/ml) and kanamycin (300 to 500 μg/ml) were added to the sterile growth media.

Deletion of SMU.152.

SMU.152 was deleted by fusion PCR as previously described (19). Briefly, ∼1.0-kb upstream and downstream regions were amplified with the primer set SMU.NLMA-NESTED-F/SMU.152-5′-FUSION-R and SMU.152-3′-FUSION-F/SMU.152-3′-R using S. mutans UA159 chromosomal DNA as a template. A Km-resistant cassette was amplified from pIB-D38 with the NCOI-KAN-F and PSTI-KAN-R primers, and overlapping fusion PCR was carried out with equal amounts of each PCR products with the primers SMU.NLMA-NESTED-F/SMU.152-3′-R. The amplified products were purified and transformed into S. mutans UA159 as previously described (6). Deletion of the SMU.152 locus was verified by PCR analysis.

Deletion of SMU.1909 to SMU.1913.

The region spanning from SMU.1909 to SMU.1913 was deleted by fusion PCR (19) as described above with the primer sets SMU.1908-F/EcoRV-SMU.1909-5′-Fusion and EcoRV-SMU.1913-3′-Fusion/SMU.1915-R using S. mutans UA159 chromosomal DNA as a template for generating upstream and downstream fragments. These two fragments were then digested with EcoRV, followed by ligation and cloning into the pGEM-T Easy cloning vector (Promega) to create pIB-D22. A Km-resistant cassette with flanking modified loxP sites was amplified from pUC4ΩKm by using the primers lox71-Km-F and lox66-Km-R and cloned into the EcoRV-digested site to generate pIB-D23. Plasmid pIB-D23 was linearized with EcoRI and transformed into UA159 and IBS-D3 to generate IBS-D6 and IBS-D8, respectively. Positive clones were verified by PCR analysis.

Construction of triple (multiple) gene knockout mutant.

Genomic DNA was isolated from ΔSMU.925-Em (35) and digested with EcoRV. Digested DNA products were then transformed into IBS-D8 and selected on THY medium plates containing Em; one such transformant was named IBS-D58.

Deletion of PTS locus.

PTS locus containing the genes SGO.1679 to SGO.1682 was deleted by fusion PCR as previously described (19). At first, ∼1.0 kb upstream and downstream regions were amplified with the primer set SGO-PTS-F/SGO-PTS-5′-FUSION-R and SGO-PTS-3′-FUSION-F/SGO-PTS-3′-R (for all primers, see Table 1) using S. gordonii DL-1 chromosomal DNA as a template. A Km-resistant cassette was amplified from pIB-D38 (constructed by cloning the Km cassette into pGEMT-Ez) using the primers NCOI-KAN-F and PSTI-KAN-R, each of which has a 20-bp overlapping region with the upstream and downstream fragments, respectively. Overlapping fusion PCR was carried out using equal amounts of each PCR product with the primers SGO-PTS-F/SGO-PTS-3′-R. The amplified products were purified and transformed into S. gordonii DL-1 as previously described (6). Deletion of the PTS locus was verified by PCR analysis.

Table 1.

Oligonucleotides used in this study

Primer	Sequence (5′–3′)	Purpose
BAMH-SMU.152-F	GCGGGATCCGGATCTGTAGTTTTTCCACAC	Cloning of SMU.152
XHO-SMU.NLM-R	GCGCTCGAG GACAACTATTAGGCTTCGGCCTTAGC	Cloning of SMU.152
XHO-Δ3′-SMU0.152-R	GCGCTCGAGTTTAAAAGATGAGTAAGATGGCAGC	Truncation of SMU.152
XHO-ΔRRSKNK-152-R	GCGCTCGAGTTTTATTTTAGCTCTTTATAGAGTTCAACAGC	Truncation of SMU.152
XHO-ΔKNK-152-R	GCGCTCGAGTTTAGCTTCTTCTTTTTAGCTCTTTATAG	Truncation of SMU.152
XHO-ΔKRRSKNK-152-R	GCGCTCGAGTTTTATAGCTCTTTATAGAGTTCAACAGC	Truncation of SMU.152
NCOI-KAN-F	CTCCCGGCCGCCATGGCGGCCGC	Amplification of Km
PSTI-KAN-R	GGTCGACCTGCAGGCGGCCGCG	Amplification of Km
BAMH-SMU.925-F	GCGGGATCCGACGCTGCCCTTGCAGCAGTG	Cloning of SMU.925
XHO-SMU.925-R	GCGCTCGAGGCTGCCATAACCACAAAAAGCC	Cloning of SMU.925
BAMH-SMU.1909	GCGGGATCCGTCCTTGGTTCGTTAATAGATTGGG	Cloning of SMU.1909
PST-SMU.1909-R	GCGCTGCAGGCCTGAAAGTAAAAACTGTTTAGCGTC	Cloning of SMU.1909
XHO-SMU.1909+SKNK-R	GCGCTCGAGTTTACTTATTCTTGCTGCGTCTCTTATGCAGTTTAATGAG	Cloning of SMU.1909+SKNK
XHO-SMU.925+SKNK-R	GCGCTCGAGTTTACTTATTCTTGCTGCGTTTTCTAAAAAAGCTGG	Cloning of SMU.925+SKNK
lox71-Km-F	CGATAACTTCGTATAATGTATGCTATACGAAGTTATGAGGATGAAGAGGATGAGGAGGCAG	Cloning of Km cassette
lox66-Km-R	CGATAACTTCGTATAGCATACATTATACGAAGTTATGCTTTTTAGACATCTAAATCTAGG	Cloning of Km cassette
SGO-PTS-F	CCGCCAGATAGATTGATAAAGTCC	Fusion PCR for ΔPTS
SGO-PTS-5′-FUSION-R	GCCGCCATGGCGGCCGGGAGGTTTGACCAAACTTTAAGAAAAG	Fusion PCR for ΔPTS
SGO-PTS-3′-FUSION-F	CGCGGCCGCCTGCAGGTCGACCGGCAAATTTTTTAAAAACCTATTTGC	Fusion PCR for ΔPTS
SGO-PTS-3′-R	GCTCATAGCTGTGAAGTAGATTGG	Fusion PCR for ΔPTS
SMU.NLMA-NESTED-F	GGCGTGACCATTCATGAAGCGAATGCC	Deletion of SMU.152
SMU.152-5′-FUSION-R	GCCGCCATGGCGGCCGGGAGCCTCTCTTGAAATTCGACATTA	Deletion of SMU.152
SMU.152-3′-FUSION-F	CGCGGCCGCCTGCAGGTCGACCGCTAAGGCCGAAGCCTAATAGTTG	Deletion of SMU.152
SMU.152-3′-R	CTGCGTTACATATGAACATAGACC	Deletion of SMU.152
SMU.1908-F	GTACTTATCCTTATTTTGTTTTTATCTTGCTG	Deletion of SMU.1909 to SMU.1913
EcoRV-SMU.1909-5′-Fusion	GCGGATATCGAACTCATTAAACTGCATAAGAGACGC	Deletion of SMU.1909 to SMU.1913
EcoRV-SMU.1913-3′-Fusion	GCGGATATCGCTGCGCTGTTAAAAGTCTTAATAACC	Deletion of SMU.1909 to SMU.1913
SMU.1915-R	CGGAAAAATGTTGATAGGCTTCCG	Deletion of SMU.1909 to SMU.1913

Construction of plasmids for complementation.

Plasmid pIB184-Km was used as vector for the cloning (19). To clone SMU.152, fragment containing the SMU.152 gene and the promoter region was amplified from the genomic DNA of UA159 using the primers BAMH-SMU.152-F and XHO-SMU.NLM-R. The amplified fragment was digested with BamHI and XhoI and cloned into BamHI-XhoI-digested pIB184-Km to generate pIB-D8. SMU.925 and SMU.1909 were also cloned using the similar approach with the primer set BAM-SMU.925-F/XHO-SMU.925-R and BAMH-SMU.1909/PST-SMU.1909-R, respectively (for all of the plasmids, see Table-2). Complementing plasmids were verified with restriction digestion and sequencing. For addition of SKNK-residues, the corresponding nucleotide sequences were added into the primers, and used for amplification of SMU.925 and SMU.1909 as described above. The amplified fragments were cloned into BamHI-XhoI-digested pIB184-Km as described above.

Different versions of truncated SMU.152 were generated using the common forward primer BAMH-SMU.152-F and various reverse primers carrying desired truncations (Table 1). The amplified fragments were digested with BamHI and XhoI, and cloned into BamHI-XhoI-digested pIB184-Km to generate respective plasmids (Table 2). Insertion of the cloned fragments was verified by restriction digestion and sequencing.

Table 2.

Strains and plasmids used in this study

Strain or plasmid	Construction and/or description	Source or reference
Streptococcus strains
S. mutans UA159	Wild type
S. gordonii DL-1	Wild type
S. constellatus	Wild type
S. mutans mutants
ΔSMU.1914 (ΔnlmC)	C. Levesque lab, University of Toronto	35
ΔSMU.925	C. Levesque lab, University of Toronto	35
IBS-D3	Deletion of nlmAB operon	19
IBS-D6	ΔSMU.1909-SMU.1913	This study
IBS-D8	ΔnlmAB ΔSMU.152, ΔSMU.1909-SMU.1913	This study
IBS-D13	ΔnlmAB ΔnlmC	19
IBS-D15	ΔPTS::loxP-Km of S. gordonii DL-1	This study
IBS-D16	ΔSMU.152::loxP-Km of UA159	This study
IBS-D58	IBS-D8; ΔSMU.925	This study
Plasmids
pIB184-Km	E. coli-Streptococcus shuttle plasmid	7
pIB-D8	pIB184-Km::SMU.152	This study
pIB-D22	pGEMT-Ez::SMU.1909-SMU.1913	This study
pIB-D23	pIB-D22::loxP-Km	This study
pIB-D33	pIB184-Km::SMU.1909	This study
pIB-D34	pIB184-Km::SMU.925	This study
pIB-D35	pIB184-Km::SMU.152 (1-120 amino acids)	This study
pIB-D36	pIB184-Km::SMU152ΔKRRSKNK	This study
pIB-D37	pIB184-Km::SMU.152ΔRRSKNK	This study
pIB-D38	pGEMT-Ez::loxP-Km	This study
pIB-D40	pIB184-Km::SMU.152ΔKNK	This study
pIB-D72	pIB184-Km::SMU.1909+SKNK	This study
pIB-D75	pIB184-Km::SMU.925+SKNK	This study

Deferred antagonism (bacteriocin) assay.

Bacteriocin assay was carried out by stabbing the isolated colonies into THY agar plates with a toothpick and grown overnight (∼18 h) at 37°C under microaerophilic conditions. Indicator strains were grown overnight in THY broth, and 0.4 ml of the culture was mixed with 10 ml of soft agar and overlaid on agar plates that were stabbed with the tester strains. Overlaid plates were then incubated overnight under same conditions and the diameter of the zones of inhibition around the mutacin producing strains was measured.

RESULTS

SMU.152 can confer immunity against mutacin IV.

SMU.152 encodes a hypothetical protein that is topologically linked to nlmAB operon, which encodes the structural genes for mutacin IV. An intergenic sequence (IGS) of 193 bp is present in between the stop codon of nlmB gene and the initiation codon of SMU.152. A schematic diagram of the SMU.152 locus and the upstream nlmAB genes is shown in Fig. 1A. Sequence analysis, confirmed using BPROM online software (prediction of bacterial promoters [Softberry]), indicates the presence of putative −35 (TTAGAA) and −10 (TATACT) box motifs 78 and 55 bp, respectively, upstream of the putative start codon with a putative ribosomal binding site (AGAGG) 11 bp upstream of the start codon. A putative rho-independent terminator sequence (AATTATGGAGGAGGA) was also identified at a position 66 bp downstream of SMU.152 by the TransTermHP (http://transterm.cbcb.umd.edu [20]) program. Thus, it appears that the SMU.152 is a single gene standalone operon.

Fig 1.

Mutacin IV and its immunity protein. (A) Genetic organization of nlmAB and SMU.152. The intergenic sequence along with the putative promoter region is shown. Coexpression of SMU.152 with nlmAB genes is verified by linkage PCR. (B) Predicted secondary structure of SMU.152 determined by HMMTOP transmembrane topology prediction software (http://www.enzim.hu/hmmtop). Abbreviations: i, inside the cell; H, helical (within membrane); o, outside of the membrane. (C) Predicted three-dimensional structure of SMU.152 as determined by MUlti-Sources ThreadER (44).

We analyzed the primary amino acid sequence of SMU.152 by several topology-predicting algorithms (TopPred [43] and HMMTOP [40]) and found that SMU.152 contains at least four transmembrane helices (Fig. 1B). The SMU.152 primary sequence was also subjected to a protein threading algorithm (MUlti-Sources ThreadER [44]) to predict the three-dimensional structure of SMU.152; the predicted structure is shown in Fig. 1C. From the structure, it appears that the last few residues are not part of any transmembrane helix and are disorganized.

To evaluate the function of SMU.152 as an immunity protein, we cloned the entire coding region and the promoter region of SMU.152 into a shuttle plasmid and introduced the plasmid into S. gordonii, which was used as the indicator strain. As shown in Fig. 2A, heterologous expression of SMU.152 in S. gordonii confers protection against mutacins secreted by S. mutans UA159. This result suggests that SMU.152 alone can give protection against mutacin IV and therefore acts as an immunity protein. We also overexpressed SMU.152 in S. constellatus, a strain that is susceptible to mutacin IV (19). We found that as in S. gordonii, the S. constellatus strain containing SMU.152 is also refractory to lysis by mutacin (Fig. 2A and B).

Fig 2.

Immunity activity of SMU.152 in S. gordonii. S. mutans UA159, and an nlmAB strain (IBS-D3) were stabbed into THY agar and incubated overnight at 37°C under microaerophilic conditions. The plates were overlaid with soft agar containing the indicator strains. The zone of inhibition of the indicator strains was evaluated after overnight incubation. The indicator strains are S. gordonii DL-1 (A) and S. constellatus (B) containing the vector (pIB184-Km) or the vector with SMU.152 (pSMU.152). (C) Mutacin activity of the ΔSMU.152 (IBS-D16) strain. These plates are representative of three independent assays.

To explore further, we attempted to delete the SMU.152 locus from UA159, leaving the nlmAB loci intact. To our surprise, we were able to construct an SMU.152 deletion strain (IBS-D16) and found no significant difference in growth in liquid medium between IBS-D16 and UA159 (data not shown). We then tested whether the IBS-D16 strain produces mutacins; as shown in Fig. 2C, IBS-D16 produces mutacin, and there was no difference in the level of production between UA159 and IBSD16. We also used IBS-D16 as an indicator strain in a deferred antagonism assay against UA159 and found that IBS-D16 is not susceptible to mutacins produced by UA159 (data not shown). Therefore, our data suggest that UA159 probably encodes other immunity proteins that confer protection.

According to the KEGG database, three other SMU.152 paralogous genes are encoded by S. mutans UA159. One of them, SMU.925, which is also known as CipI, was previously shown to function as an immunity protein against mutacin V (encoded by nlmC) (35). Two other putative immunity proteins, SMU.1909 and SMU.1913, are located near putative bacteriocin-like proteins (Blp) such as SMU.1906 and SMU.1914. Since SMU.152 shows sequence homology with SMU.925 as well as with SMU.1909 and SMU.1913, we wanted to test the possible role of these proteins for immunity to mutacin IV. We constructed various deletion derivatives of these putative immunity genes and used the strains as indicators against UA159. When we used as an indicator, a strain in which SMU.925 was deleted (ΔSMU.925) or a strain in which both SMU.1909 and SMU.1913 were deleted (ΔSMU.1909-SMU.1913), we did not observe any significant changes in the sensitivity to mutacin IV or V (data not shown). We then constructed a multiple deletion strain in the ΔnlmAB background, where we knocked out SMU.152, SMU.925, and SMU.1909 to SMU.1913. When we tested this IBS-D58 strain as an indicator, to our surprise we found that the strain became sensitive to its own mutacins (Fig. 3). However, the zone of lysis was not as large as the zones produced against S. gordonii or other indicator bacteria. Taken together, our results suggest that SMU.152 may interact with other immunity proteins to confer complete protection against its own mutacins.

Fig 3.

Immunity activity of different mutants. UA159, ΔnlmAB, ΔnlmC, and ΔnlmABC strains were stabbed into THY agar, and the mutacin activity was measured against IBS-D16 (ΔSMU.152) and IBS-D58 (ΔnlmAB, ΔSMU.152, ΔSMU.1909-SMU.1913, and ΔSMU.925). These plates are representative of three independent assays.

C-terminal charged residues of SMU.152 are required for optimum activity.

Since SMU.152 confers immunity against mutacins, and because the predicted three-dimensional structure indicates a disordered coil at the C terminus, we wanted to evaluate the role of the C terminus in mutacin immunity. We made sequential deletions of 19, 7, 6, and 3 residues from the C-terminal region and tested for immunity activity. As shown in Fig. 4, deletion of as few as six residues from the C-terminal region caused drastic reduction in the immunity activity. Therefore, it appears that charged residues KRRSKNK at the C-terminal regions are required for optimum immunity activity against mutacins.

Fig 4.

Immunity activity of truncated SMU.152. Various C-terminal deletions of SMU.152 were created and tested for immunity activity in S. gordonii as described in Fig. 2. The data shown (in millimeters) are means ± the standard deviations (error bars) of at least three independent experiments. A Student t test was used to calculate the significance of the difference between the mean diameters of the zones of lysis. Asterisks indicate the P values that are <0.001.

SMU.1909 with SKNK residues can confer immunity against mutacin IV.

As mentioned above, SMU1909 shows significant homology to SMU.152 and is annotated as a paralogous gene for SMU.152. When we performed a sequence alignment (Fig. 5A), we found that SMU.1909 protein shares 49% similarity (including 30% identity) with SMU.152. Most importantly, we found that SMU.1909 lacks the last four residues of the KRRSKNK sequence. To verify whether SMU.1909 can confer immunity against mutacin IV, we overexpressed SMU.1909 and SMU.1909 carrying SKNK residues in S. gordonii. As shown in Fig. 5B, expression of SMU.1909 does not provide protection (compare the zone of lysis for the top two panels). However, when we added SKNK sequence to SMU.1909, the zone of lysis was reduced to approximately half compared to the vector control. Thus, it appears that whereas SMU.1909 alone does not confer immunity against mutacin IV, when we restored the complete KRRSKNK sequence that is present in SMU.152, we observed significant amount of protection against mutacin IV conferred by the modified SMU.1909.

Fig 5.

Immunity activity of SMU.1909. (A) Sequence alignment between SMU.152 and SMU.1909. An asterisk (*) indicates positions that have a single, fully conserved residue; a colon (:) indicates conservation between groups of strongly similar properties; and a period (.) indicates conservation between groups of weakly similar properties. (B) S. gordonii strains containing SMU.1909 alone or SMU.1909 with four additional residues (SKNK) were tested for sensitivity against mutacin IV produced by UA159. Assays were repeated at least three times, and a representative plate is shown. A Student t test was used to calculate the significance of the difference between the mean diameters of the zones of lysis (P < 0.001).

SMU.925 with SKNK also can confer immunity to mutacin IV.

As mentioned above, SMU.925 also shows significant similarity with SMU.152. Since SMU.925 also shares sequence homology with SMU.152 (Fig. 6A), we wanted to test whether overexpression of SMU.925 in S. gordonii can confer protection against mutacin IV. As shown in Fig. 6B, over expression of SMU.925 did not produce noticeable difference in the zone of lysis (compare the top panel with the middle panel). Like SMU.1909, SMU.925 also lacks the last four charged amino acids that are present in SMU.152. We added these residues (SKNK) to SMU.925 and tested sensitivity against mutacin IV. We observed a slight but reproducible decrease in the zone of lysis produced by wild-type UA159 (Fig. 6B, bottom panel). To further confirm that SMU.925+SKNK can confer immunity against mutacin IV, we tested against an ΔnlmC strain that predominantly produces mutacin IV (19). As observed before, the zone of lysis produced by the ΔnlmC strain in S. gordonii is smaller than the zone produced by the wild-type strain (19). However, when we overexpressed SMU.925+SKNK in S. gordonii, the zone of lysis was significantly smaller (∼25% reduction) compared to the vector control strain. Thus, it appears that addition of SKNK to SMU.925 can confer some level of immunity against mutacin IV.

Fig 6.

Immunity activity of SMU.925. (A) Sequence alignment between SMU.152 and SMU.925. The symbols are described in Fig. 5A. (B) A deferred antagonism assay was carried out with S. gordonii DL-1 strain as described in Fig. 5B, except that the indicator strain also contains the vector plasmid to facilitate selection during antagonism assay. Assays were repeated at least three times, and a representative plate is shown. A Student t test was used to calculate the significance of the difference between the mean diameters of the zones of lysis, and the P values are shown.

Man-PTS system is required for optimum mutacin IV activity.

The Man-PTS, which is a major sugar uptake system in Firmicutes, was recently shown to act as a receptor for class IIa bacteriocins and for lactococcus A, a class IIc bacteriocin (12, 21, 22). However, these two classes differ from class IIb bacteriocins since they are single-peptide bacteriocins. To verify whether Man-PTS system plays a role as a receptor for mutacin IV, we deleted the entire Man-PTS system from S. gordonii and subjected the deletion strain in a deferred antagonism assay for mutacin IV sensitivity. As shown in Fig. 7, the zone of lysis was ca. 30% smaller in the Man-PTS deleted strain compared to the wild-type S. gordonii strain. Thus, Man-PTS system appears to play a role in mutacin IV sensitivity. That we observed some degree of lysis in the Man-PTS deleted strain suggests that the Man-PTS is not the only complex that acts as receptor for mutacin IV in S. gordonii.

Fig 7.

Mannose-PTS is required for optimum sensitivity. Assays were done as described in Fig. 2. UA159 was stabbed into a THY agar plate. The indicator strains are wild-type S. gordonii DL-1 (SGO) and a mutant DL-1 without a functional Man-PTS system (ΔPTS, IBS-D15). Assays were repeated at least three times, and a representative plate is shown. A Student t test was used to calculate the significance of the difference between the mean diameters of the zones of lysis (P < 0.001).

DISCUSSION

Bacteriocins that belong to class II generally kill target cells either by pore formation or by interfering with the integrity of the cell membrane; however, the exact molecular pathways by which the bacteriocins exert their effect on membrane are unknown (13). Since each bacteriocin displays a specific inhibitory spectrum, this indicates that individual bacteriocins might recognize a particular receptor on the target cells. Some immunity proteins, specifically those that are related to class IIa and IIc bacteriocins, have been shown to bind directly to these cell surface receptors and block pore formation (12). However, no immunity mechanism is known for any class IIb two peptide bacteriocins, except for the Abi family proteins that presumably degrade the bacteriocins. In the present study, we characterized a self-immunity protein for mutacin IV, a two-peptide bacteriocin.

The genes that encode self-immunity proteins for class IIb are often located just downstream of the structural genes for bacteriocin. Because of this genetic organization, putative immunity proteins for most of the bacteriocins can be readily identified. However, their function as an immunity protein has been demonstrated for only few of them (33). SMU.152 is also located downstream of nlmAB operon and, therefore, it was an appropriate candidate to test for immunity activity. Indeed, our results indicate that SMU.152 when expressed in heterologous sensitive hosts can confer complete protection against mutacin IV, suggesting that SMU.152 alone is sufficient to counteract mutacin IV activity. However, surprisingly, we found that when SMU.152 was deleted from S. mutans, the deletion strain was still refractory to mutacin IV activity. Several possibilities can explain why the SMU.152-deficient strain is resistant to mutacin killing. For example, there may be other immunity proteins in S. mutans to offer cross-protection when the cognate immunity protein for a specific mutacin is absent. However, when we tested two putative immunity proteins (SMU.1909 and SMU.925) in S. gordonii, including the one (SMU.925) that confers protection against mutacin V (35), we did not observe any protection. Although this does not rule out that these two proteins cannot function as an immunity protein in the native host and may require a cellular component as suggested for LcnG, the lactococcal G immunity protein (33). Moreover, genome analysis indicates that S. mutans encodes several other proteins with similar structural features and sizes, and some of these proteins could function as immunity proteins. Furthermore, this possibility has been corroborated by the findings that S. mutans OMZ175, which does not carry the nlmAB operon, is insensitive to mutacin IV (data not shown). Moreover, deletion of SMU.1909 in an nlmAB-deficient strain (19) and OMZ175 (which lacks nlmAB locus) do not show sensitivity against mutacin IV (data not shown). Our data also suggest that SMU.152 may interact with other putative immunity proteins to confer full protection in the native host since deletion of multiple putative immunity proteins leads to sensitivity.

Our data strongly indicate that SMU.152 is sufficient to provide immunity to sensitive strains. We also showed that the C-terminal seven charged residues are important for the immunity activity. Theoretical structure prediction suggested that SMU.152 contains four helical domains and the last 10 residues are highly disorganized. Deletion of last seven amino acids completely abolished the immunity function of this protein. Thus, these charged amino acids are important for the activity. We speculate that this region of the protein directly interacts with the mutacin peptides and therefore provides specificity. A recent study on lactococcal G suggests that the N-terminal 13 residues of one peptide (Lcnα) and 11 residues at the C-terminal half of the other peptide (Lcnβ) are involved in the recognition by the immunity protein (33). Since mutacin IV is highly homologous to lactococcin G, we speculate a similar situation where the N-terminal portion of NlmA and the C-terminal portion of NlmB will interact with SMU.152. It is possible that this two-peptide mutacin IV might interact with a membrane-integrated protein that functions as the receptor and induce membrane leakage. Binding of SMU.152 to the mutacin IV-receptor complex blocks the membrane leakage. This mechanism is similar to what has been demonstrated for class IIa pediocin-like bacteriocin involving the Man-PTS system as the receptor (12). It is important to point out that we also identified Man-PTS as one of the putative receptors for mutacin IV activity since deletion of the Man-PTS rendered the S. gordonii strain less sensitive to mutacin IV. That we did not observe complete protection against mutacin IV when Man-PTS was deleted indicates that there may be other cell surface proteins in S. gordonii that function as mutacin IV receptor.

Spontaneous resistance can occur for some class II bacteriocins when the sensitive cells are continuously exposed to bacteriocins. The frequency of such occurrence depends on the type of bacteriocins, as well as on the type of sensitive cell. Listeria monocytogenes can achieve resistance frequency as high as 10⁻⁴ when exposed to class IIa type bacteriocins, including leucocin A and pediocins (15). The mechanisms of resistance generation can vary depending on the nature of the bacteriocin, and it has been extensively studied for class IIa type bacteriocins. In most cases, the expression of Man-PTS is downregulated in resistant mutants (15, 21, 34, 39). However, Man-PTS independent generation of resistance cells has also been reported in L. monocytogenes where changes in the cell envelope have been found to be associated with the resistant mutants (41, 42). Thus far, no spontaneous resistance mutant has been isolated for mutacin IV, and we speculate that the frequency of resistance occurrence would be very low.

Mutacin IV produced by S. mutans can inhibit the growth of all groups of streptococci except for mutans group streptococci, including S. criceti, S. downei, S. ratti, and S. sobrinus (19). The exact mechanisms by which these mutans streptococci are immune to mutacin IV are not known. It is possible that the entire mutans streptococcus group lacks a functional receptor for mutacin IV and possibly for other mutacins as well. Alternatively, these mutans streptococci encode the immunity genes without the corresponding mutacin genes, a phenomenon known as the immune mimicking mechanism, and this has been shown for sakacin and pediocin immunity (14, 26). Mutans streptococci may also encode transporters to pump peptides out from the cell envelope, thereby conferring resistance against mutacins, as has been shown for some lantibiotics (9, 25). Further investigation, such as comparative genome analysis of the sensitive and resistant streptococci, may reveal additional mechanisms of immunity against mutacins secreted by S. mutans.