INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Several bacterial species produce peptidoglycan hydrolases that lyse target cells (4, 8, 10, 33, 43). The physiological functions of these enzymes remain largely unknown. The suggestion that bacteria use them to gain a competitive advantage in environments where nutrients are limited has been made previously (49). They may also allow producer organisms to establish an ecological niche within a competitive environment (49).

Millericin B, a peptidoglycan hydrolase produced by Streptococcus milleri NMSCC061, was purified and partially characterized as an endopeptidase that cleaves the interpeptide cross bridge and the stem peptide of susceptible peptidoglycan (6). This suggests that millericin B has activity similar to that of lysostaphin and ALE-1, both of which are glycylglycine endopeptidases that are produced by Staphylococcus simulans bv. staphylolyticus and Staphylococcus capitis EPK1, respectively (43, 47), as well as zoocin A, an endopeptidase produced by Streptococcus equi subsp. zooepidemicus 4881 (4).

Molecular cloning of the ale-1 gene showed that the primary structure of the mature ALE-1 is very similar to that of the proenzyme form of lysostaphin. The genes for lysostaphin (end) in S. simulans bv. staphylolyticus and for ALE-1 (ale-1) in S. capitis EPK1 were found to be located on large plasmids (20, 22, 35). This was in contrast to a previous finding where the gene for lysostaphin was found to be located on the chromosome of S. simulans bv. staphylolyticus (24).

S. simulans bv. staphylolyticus is resistant to lysis by lysostaphin (37), and this resistance is conferred by modifying the amino acid composition of interpeptide chains in cell wall peptidoglycan by increasing the serine content and decreasing the glycine content (11). The genetic elements for production and self-protection (immunity) of both lysostaphin and ALE-1 have been characterized (11, 22, 47, 50). The gene involved in the modification of peptidoglycan (epr, for endopeptidase resistance) was shown to be located on a large plasmid, pACK1, together with the gene (end) which encodes the endopeptidase (22). Staphylococcus aureus cells transformed with a plasmid containing the 8.4-kb DNA fragment from pACK1 produced lysostaphin and were resistant to lysis by lysostaphin, which indicated that the DNA fragment contained both epr and end (11).

The objective of this study was to determine whether the genes for millericin B and millericin B host immunity in S. milleri NMSCC 061 are similar in organization to the genes for lysostaphin and lysostaphin host immunity in S. simulans bv. staphylolyticus and the genes for ALE-1 and ALE-1 host immunity in S. capitis EPK1 (46, 47, 50). Here we describe the sequencing of the millericin B gene (milB) and the identification and sequencing of three additional genes in S. milleri NMSCC 061 that appear to be associated with millericin B host self- protection and its export. One of these genes (milF, for millericin B immunity factor), which is similar in sequence to genes that have been implicated in the synthesis of peptidoglycan cross bridges in staphylococci and streptococci (5, 12, 45, 47, 50), appears to specify the incorporation of a leucine in place of a threonine residue in the cross bridge of S. milleri NMSCC 061. A second gene is a putative tRNAleu gene, which is presumably involved in the incorporation of the leucine in the cross bridges of the peptidoglycan. The product of the third gene (milT) is homologous to a class of ATP-dependent transport proteins and may be necessary for the export of millericin B.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains and culture conditions. The millericin B producer strain S. milleri NMSCC 061 and the nonproducer S. milleri NMSCC 051 were grown in tryptone soy broth or tryptone soy agar plates containing 1.4% agar per liter. All working cultures were subcultured every second week and stored at 4°C until needed. Stock cultures were maintained in tryptone soy broth with 30% glycerol at 70°C.

Purification of peptidoglycan. A 100-ml culture (using a 1% inoculum from a culture maintained on cell wall minimal medium [28.5 mM glucose, 1 mM DL-glutamic acid, 0.75 mM alanine, 0.2 mM lysine, 0.17 mM uracil, 0.125 mM glycerol, 1 mM MgCl₂, 0.1 mM MnCl₂, 3 µM thiamine, and 8.2 µM nicotinamide in 80 mM potassium phosphate buffer, pH 6.8, filter sterilized.]) of either S. milleri NMSCC 061 or S. milleri NMSCC 051 was grown in cell wall minimal medium overnight or until the optical density at 600 nm reached 1.0. After centrifugation (2,000 × g, 4°C) cells were washed twice with saline, once with water, and three times with acetone. The sediment was dried at 37°C, and 2.5 g of the dried powder was suspended in 25 ml of ice-cold distilled water and mixed with 25 ml of 0.1- to 0.15-mm-diameter glass beads in a Virsonic Sonicator until total disruption of cells was achieved (5 to 15 min). The glass beads and unbroken cells were removed after centrifugation for 10 min at 1,500 × g followed by a second centrifugation at 6,500 × g to sediment the cell walls. The pellet was washed three times with distilled water. Hydrolysis of the cellular macromolecules trapped in the cell wall was achieved by incubation with RNase A (100 µg/ml) and DNase (50 µg/ml; Roche Biochemicals) for 18 h at 37°C followed by trypsin (200 µg/ml; Sigma) for another 18 h at 37°C with buffer conditions according to the manufacturer's instructions. The walls were collected after centrifugation at 6,500 × g for 30 min and washed four times with water. Teichoic acids were extracted twice using 250 ml of 5% trichloroacetic acid for 18 h at 22°C. Peptidoglycan was collected by centrifugation at 6,500 × g for 30 min, washed three times with distilled water and three times with acetone, dried, and stored frozen until needed.

Amino acid analysis of peptidoglycan. Amino acid analysis was performed by conventional reverse-phase chromatography and O-phthaldialdehyde (OPA) derivatization. The OPA reagent was kept as a stock solution of 56 mg of OPA dissolved in 1 ml of methanol and 10 µl of 2-mercaptoethanol. This stock solution was stored at 4°C and diluted 1:10 with sample buffer (0.4 M Na₃BO₃ [pH 9.5]) for use. This dilution was freshly prepared daily. Standard amino acids were dissolved in 0.1 N HCl and peptidoglycan hydrolysates in sample buffer. Portions containing 20 pmol of each amino acid were subjected to derivatization by mixing 1:9 (vol/vol) with OPA reagent. A 20-µl portion was injected onto an Microsorb C₁₈-RF column (particle size, 5 µm; Rainin Instruments Co., Inc., Woburn, Mass.) after a 60-s derivatization period. OPA-derivatized amino acids were separated on a Perkin-Elmer gradient system with a methanol-3% tetrahydrofuran (by volume) gradient in 50 mM Na₂PO₄ (pH 6.5) (40). The flow rate was kept at 1 ml/min.

Bactericidal action of millericin B on S. milleri NMSCC 051 and cell wall variants of S. milleri NMSCC 051 cells. A suspension of log-phase cells of S. milleri NMSCC 051 or the S. milleri NMSCC 051 Thr-to-Leu mutant was centrifuged, washed twice and resuspended in 20 mM phosphate buffer (pH 7.0) to an optical density at 600 nm of 0.5. Aliquots (0.1 ml) taken at 10-min intervals of suitable 10-fold dilutions were plated on tryptone soy agar plates, and the number of CFU was recorded after 18 h of growth.

Production of cell wall cross bridge variant. To change the amino acid composition of the interpeptide cross bridges of S. milleri NMSCC 051, cells were grown for 20 successive subcultures in 10-ml cultures of cell wall minimal medium (54). Leucine was added to a concentration of 1 mM. The peptidoglycan composition of cells was checked by analyzing the amino acids present in the peptide moiety using OPA derivatization.

Cleavage of peptidoglycan and liberation of free amino groups. Peptidoglycan (1 mg/ml) was incubated with millericin B (10 µg/ml) and aliquots (10 to 50 nmol, estimated from a standard curve) taken at different time intervals to determine the liberation of free amino groups (6). Flourodinitrobenzene (FDNB) reagent (130 µl in 10 ml of 100% ethanol) was added to each aliquot, mixed, and incubated at 60°C for 30 min. After acidification with concentrated HCl (50 µl) the DNP derivatives were detected by spectrophotometric adsorption at 405 nm. Reactions containing enzyme and FDNB reagent and those containing peptidoglycan and FDNB reagent, respectively, were used as controls.

Well diffusion assay. Purified peptidoglycan (1%) was incorporated into agar, and an aliquot of 10 µl of a 5-mg/ml solution of millericin B was added to the wells in the agar. Plates were incubated at 37°C for 18 h and checked for zones of clearing.

Purification of chromosomal DNA. Genomic DNA was isolated from overnight cultures of S. milleri NMSCC 061 using the NucleoSpin C + T DNA extraction Kit (Macherey-Nagel, Düren, Germany) with modifications. The buffer used to lyse the cells contained lysozyme (4 mg/ml), 20 mM Tris-Cl, 2 mM EDTA, and 1% Triton, pH 8.0. Cells were incubated for 1 h with lysis buffer at 37°C to ensure complete lysis. Purified genomic DNA was eluted from the NucleoSpin column with 10 mM Tris-HCl, pH 8.5, and used directly for amplification or for other enzymatic reactions.

Generation of internal protein fragments for amino acid sequencing. Purified millericin B was cleaved with cyanogen bromide to generate internal fragments. Millericin B (5 mg/ml) was solubilized in 50 µl of 70% formic acid. A crystal of cyanogen bromide was added, and the reaction tube was flushed with nitrogen upon closing. After 12 h of incubation at room temperature in the dark, the formic acid was evaporated under a vacuum. The digest was solubilized in a solution containing 6 M guanidine-HCl, 0.1 M Tris (pH 8.5), and 0.1 M dithiothreitol and resolved using reverse-phase high-performance liquid chromatography. Peaks were collected separately, and N-terminal amino acid sequence was obtained using an API 491 Procise automated sequencer (Applied Biosystems, Perkin-Elmer).

Construction and design of primers for PCR and DNA sequencing. Primers were constructed using reverse genetics from the amino acid sequences obtained from purified millericin B (6). The forward primer, 5'GAAAATGATTTTAGTCTAGCAATG3' (Mill-1), was designed from the N-terminal amino acid sequence of purified millericin B (6). The reverse primer, 5'CTATTTACAGCCGTAAGGGCC3' (Mill-2), was designed from the amino acid sequence obtained from an internal protein fragment (6). A third primer, Mill-3, was designed from the sequence generated with Mill-2 and Mill-1. Primers Mill-1 and Mill-2 were synthesized by the Department of Biochemistry, University of Cape Town. Primer Mill-3 used for sequencing was synthesized by Sequiserve, Verstetten, Germany. DNA sequence was determined by the dideoxy chain-termination method (42) with Thermo Sequenase (Amersham International, Little Chalfont, United Kingdom) by using an automated DNA sequencing system (ALFred; Pharmacia) at Sequiserve. Both strands were sequenced by using the PCR primers Mill-1 and Mill-2. Primer Mill-3 was used to sequence the internal region of the 1.8-kb fragment. The 3.2-kb fragment was sequenced using the pUC universal forward and Mill-2 primers. Internal sequence was generated by using primer walking.

PCR. Fragments of genomic DNA were amplified with standard reagents from Roche Biochemicals (Mannheim, Germany), using a GeneAmp System 2400 thermocycler (Perkin-Elmer). For the generation of an internal 1.8-kb fragment (see Fig. 4) the initial amplification was carried out under low-stringency conditions followed by higher-stringency conditions according to the following profile: (i) 95°C for 5 min, 37°C for 3 min, 70°C for 3 min (3 cycles); (ii) 94°C for 1 min, 45°C for 1 min, 72°C for 3 min (30 cycles); and (iii) 72°C for 7 min. A concentration of 3.5 mM magnesium chloride (MgCl₂) yielded the brightest bands. In order to generate PCR fragments that would provide information on the sequence upstream of the millericin structural gene, genomic DNA from S. milleri NMSCC 061 partially digested with EcoRI was ligated into the EcoRI site of pUC 118. The ligation mixture was used as the template in a PCR, according to standard procedures, with Mill-2 and the pUC universal forward primers (see Fig. 4). The amplified fragments of interest were purified using either the agarase treatment method (Roche Biochemicals) or the NucleoSpin Extract Kit (Macherey-Nagel) as per manufacturers' instructions. PCR-amplified products were electrophoresed on 1.5% agarose (Whitehead Scientific, S.A.) gels. The amplified fragments were excised from low-melting-point agarose gels (Whitehead Scientific, S.A.). Gel pieces were treated with agarase (Roche Biochemicals, Germany), or gel fragments were purified directly using a NucleoSpin extraction kit (Macherey-Nagel). For agarase treatment gel pieces were incubated at 65°C for 15 min and cooled to 45°C. Typically 1 U of agarase was added for every 100 mg of agarose and incubated at 45°C for 1 to 2 h. Precipitation of the DNA was done using conventional methods (41). The purified fragments were dissolved in sterile distilled water for further use.

Nucleotide sequence accession number. The nucleotide sequence data identified in this study appears in the DDBJ, EMBL, and GenBank nucleotide sequence database under accession number AF243359.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Amino acid analysis of S. milleri NMSCC 061 (producer, resistant to lysis by millericin B) and S. milleri NMSCC 051 (sensitive). The cell wall composition of S. milleri strain NMSCC 051 was analyzed and compared to that of the millericin B producer strain S. milleri NMSCC 061. Both strains contained the usual O-acetylated N-acetylmuramic acid and N-acetylglucosamine sugar moieties in their peptidoglycan backbone. The two strains differed in the amino acid composition of their peptide moieties as determined by OPA derivatization. The resistant producer strain (NMSCC 061) had a leucine in its interpeptide cross bridge, whereas the sensitive strain (NMSCC 051) had a threonine residue in the same position.

Production of a cell wall variant. A cell wall variant of S. milleri NMSCC 051 (sensitive strain) was created in which 40 to 50% of the threonine residues were substituted with leucine residues (Fig. 1). The molar masses of each amino compound were normalized with respect to glutamic acid. The variant was less sensitive to millericin B (Fig. 2). The substitution of leucine for threonine in the cell wall variant was not able to confer total resistance to millericin B. Only 50% as many free amino groups were liberated during millericin B digestion of the peptidoglycan of the cell wall variant as were released during digestion of the peptidoglycan of the wild-type organism (Fig. 3).

View larger version (44K): [in this window] [in a new window]   FIG. 1.   The amino compound composition of peptidoglycan of three S. milleri strains. The molar masses (y axis) of each amino compound were normalized with respect to glutamic acid (not shown), which was used as a standard in the amino acid analysis assay. N-Acetylmuramic acid (mur) and N-acetylglucosamine (ngl) are also shown.

View larger version (11K): [in this window] [in a new window]   FIG. 2.   Viable cell count of cultures incubated with millericin B. Symbols: , S. milleri NMSCC 061; , S. milleri NMSCC 051; , S. milleri NMSCC 051 Thr-to-Leu mutant.

View larger version (12K): [in this window] [in a new window]   FIG. 3.   Liberation of free amino groups from peptidoglycan digested with millericin B. Purified peptidoglycan was incubated with millericin B and the products derivatized with FDNB. Symbols: , S. milleri NMSCC 051; , S. milleri NMSCC 051 Thr-to-Leu mutant; , control.

Generation of PCR fragments for sequencing. The millericin B operon (Fig. 4) nucleotide sequence (Fig. 5) was determined. The strategy focused on generating PCR fragments using reverse genetics from known amino acid sequences of purified millericin B (6). A 1.8-kb fragment was generated from primers Mill-1 and Mill-2 and sequenced. A third primer, Mill-3, derived from the DNA sequence generated with Mill-1 and Mill-2, was used to complete the region flanking the Mill-1- and Mill-2-generated sequences. A second PCR strategy using Mill-2 and the pUC universal forward primer was used (Fig. 4A). Genomic DNA from S. milleri NMSCC 061 partially digested with EcoRI was ligated into the EcoRI site of pUC 118. The ligation mixture was used as the template in a PCR, according to standard procedures, with Mill-2 and the pUC universal forward being the amplification primers.

View larger version (11K): [in this window] [in a new window]   FIG. 4.   (A) Strategy for amplification and sequencing of the millericin B operon. Mill-1 and Mill-2 primers were derived from the NH2 terminus and internal amino acid sequences of purified millericin B, respectively, and were also used as sequencing primers. Primer Mill-3 was designed using the sequence obtained from Mill-2 sequence data. The template for PCRs was generated by ligating partially digested genomic DNA into the EcoRI site of pUC 118. (B) Organization of the millericin B locus on the chromosome of S. milleri NMSCC 061. Thick arrows indicate the direction in which the genes are transcribed. milT, millericin transporter gene; tRNA-leu, cell wall-specific leucine tRNA sequence; milB, millericin B structural gene; milF, millericin immunity factor.

View larger version (140K): [in this window] [in a new window]   FIG. 5.   Nucleotide sequence and deduced amino acid sequence of the millericin B locus contained on the genome of S. milleri NMSCC 061. The predicted amino acid sequences of the ORFs are shown below the nucleotide sequence. The putative 35 and 10 promoter regions as well as the putative ribosome binding sites (RBS) are underlined. The underlined sequence is the putative tRNALeu coding region. The vertical arrow in the amino acid sequence of MilB indicates the mature processing site.

Nucleotide sequence analysis. Analysis of the nucleotide sequence (Fig. 5) revealed a polycistronic operon containing four open reading frames (ORFs). The second ORF (nucleotides 3066 to 3905) corresponds to a translation product of 277 amino acids. This predicted amino acid sequence includes a region identical to that shown to be present at the NH₂ terminus of purified millericin B (6) and indicates that the corresponding nucleotide sequence is the millericin B gene, designated milB. An NH₂-terminal extension of 18 amino acid residues contains a cluster of four positively charged residues followed by an uncharged largely hydrophobic sequence and, therefore, has the properties characteristic of a signal peptide (23). In the NH₂-terminal extension there is a GG cleavage motif at position 2 and 1, which is common to bacteriocin leader sequences. A putative promoter sequence was identified upstream of the first ORF. The sequences TTTCACA (position 231 to 237) and TATTATT (position 253 to 259) at 35 and 10, as well as their spacing (16 nucleotides), correspond closely to the sequences and spacing found in other (gram-positive) constitutive promoters (53). Protein homology searches (Blast, Swiss Prot, and EMBL databases) revealed that the deduced amino acid sequence of millericin B has striking homology with lysostaphin and ALE-1, which are glycylglycine endopeptidases produced by S. simulans bv. staphylolyticus and S. capitis, respectively. Millericin B was found to have 58% similarity to ALE-1 and 62% similarity to lysostaphin (Table 1). The N-terminal region of millericin B also shows strong homology (82%) to the N-terminal region of ALE-1 and lysostaphin.

Putative transporter gene. ORF 1 (milT) encodes a predicted protein consisting of 724 amino acid residues with a TAG stop codon (Fig. 5). Searching the protein sequence databases revealed strong similarities between the milT protein and several other proteins. The most similar proteins were members of the ATP-dependent transport proteins (13, 26), with several bacteriocin ATP-dependent transporters (19) having the greatest similarity.

The millericin B immunity factor. A gene (ORF 3) that codes for a putative millericin B host immunity factor was identified downstream from the millericin B structural gene (Fig. 4B). Figure 5 shows the nucleotide sequence of the PCR fragments sequenced. The millericin B host immunity factor (milF) starts with an ATG codon at nucleotide 3972 and ends with a TTA codon at nucleotide 4886. A Shine-Dalgarno sequence, AGAATGA, which is similar to those of streptococci (29) was observed 6 nucleotides upstream of the putative start codon. The ORF codes for a 302-amino-acid protein with a predicted molecular mass of 35.932 kDa.

Putative cell-wall-specific tRNA. A region of the nucleotide sequence located downstream of the putative milT gene codes for a putative tRNA species that shows great similarity to the tRNA^Leu sequence of streptococci and E. coli (3). This putative leucyl-tRNA could be a cell-wall-specific tRNA involved in leucine incorporation, in the interpeptide bridge during peptidoglycan synthesis. This could be equivalent to the cell wall-specific glycyl-tRNAs and seryl-tRNAs of S. aureus and S. epidermidis (27, 36, 44).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Results of this study demonstrate that the cell wall composition of S. milleri is not static but one which can be altered by the composition of the medium in which the bacteria are grown. Serial growth of the organism in a cell wall minimal medium containing an increased concentration of leucine resulted in the increased in vitro incorporation of leucine into the mucopeptide of the cell wall. As a result of leucine incorporation the variant organism S. milleri NMSCC 051 Leu-to-Thr mutant became more resistant to millericin B.

Several mechanisms have been proposed for the resistance of cell wall peptidoglycan to peptidoglycan hydrolases. For example, O-acetylation or unacetylation of amino sugars in peptidoglycan makes the cells resistant to lysozyme and other peptidoglycan hydrolases (1, 21, 31, 32, 39, 48, 51). Accessory cell wall polymers, such as lipoteichoic acid (7, 9, 16, 25, 28) teichoic acid (18, 50, 52), and teichuronic acid (38), are implicated as endogenous inhibitors of peptidoglycan hydrolases and are suggested to protect peptidoglycan.

Previous investigators reported that the amount of serine incorporated into the cross bridge of S. epidermidis is dependent on the ratio of glycyl- and seryl-tRNAs in an in vitro synthesizing system (44). A species of glycyl-tRNA (36, 44) and a species of seryl-tRNA (34) have been reported which participate in staphylococcal peptidoglycan synthesis but not in protein synthesis. Located within the millericin B operon we found a sequence that codes for a leucyl-tRNA species (Fig. 5). This species could play a role in the incorporation of leucine into the peptide cross bridge during peptidoglycan synthesis. In S. simulans bv. staphylolyticus, modification of the amino acid composition in peptidoglycan interpeptide cross bridges, as specified by the gene for lysostaphin resistance (epr or lif), was found to confer resistance to lysostaphin in these organisms (11, 50). Possibilities of involvement of additional polymers, such as teichoic acid or lipoteichoic acid, and acetyl moieties of the amino sugars in lysostaphin resistance were excluded (50).

A protein homology search revealed that the deduced amino acid sequence of the milF gene of S. milleri NMSCC 061 is similar to that of femAB of S. aureus (31, 45). milF also shares significant homology to murM (56%) and murN (48%), two proteins from Streptococcus pneumoniae responsible for interpeptide bridge formation in penicillin-resistant strains of pneumococci (15).

Detailed analysis of the peptidoglycan structure of a femA mutant revealed an accumulation of mono-glycyl-substituted muropeptide and several species of muropeptides with substitution by serine in the second or fourth positions of the interpeptide chains (12). These results suggest that the biosynthetic block in the femA mutant occurs after the addition of the first glycine residue to the interpeptide chain. Further studies suggested that femA is involved in the addition of the second and third glycines to the first glycine and that femB is involved in the addition of the fourth and fifth glycines of the interpeptide pentaglycine chain of the peptidoglycan precursor (31, 45). Our results suggest that the milF gene product is involved in adding leucine to the interpeptide cross bridge of the peptidoglycan in S. milleri NMSCC 061 and that this gene belongs to a family of genes that includes femAB, the epr genes in S. simulans bv. staphylolyticus and S. capitis EPK1, the zif (zoocin immunity factor) gene in S. equi subsp. zooepidemicus, and murMN (15) that are involved in the biosynthesis of peptidoglycan cross bridges in gram-positive cocci. Furthermore, in vivo incorporation of a leucine residue into the peptide cross bridge of S. milleri NMSCC 051 made its peptidoglycan more resistant to millericin B cleavage. Considering that the producer strain S. milleri NMSCC 061 differs from the susceptible strain, S. milleri NMSCC 051, by only a leucine residue in its peptide cross bridge, we suggest that the alteration of the amino acid composition in peptidoglycan interpeptide chains of S. milleri NMSCC 061 confers some resistance to millericin B.

The milT gene showed homology to several bacteriocin ABC transporters as well as a group of eucaryotic proteins involved in multidrug resistance. The predicted milT protein consists of an N-terminal half which is largely hydrophobic in character, while the C-terminal half is mainly hydrophobic. The similarity of MilT to members of the ATP-binding transport family (26) strongly suggests that its function includes a specific transport activity. The comparisons showed most similarity to HlyB, an E. coli membrane protein for the export of hemolysin, and competence factor ComA from S. pneumoniae (14, 29). On the basis of homology to the ComA and HlyB proteins, six potential membrane-spanning segments (30), clustered in the N-terminal half of the sequence, were identified in the predicted MilT protein (Fig. 6). In the C-terminal part of the protein, an ATP-binding motif (GMSGSGKTT) at position 519 to 527 and a region unique to proteins of the active transport group are highly conserved (Fig. 6) (12, 23). In addition, a proteolytic domain is observed in many of the ABC exporters that is responsible for cleavage of the leader peptide during export (19).

View larger version (88K): [in this window] [in a new window]   FIG. 6.   Alignment of S. milleri NMSCC 061 MilT and S. pneumoniae ComA (29) deduced protein sequences. The sequences were aligned by using the Clustal X program. The asterisks indicate identical residues, and residues with high and low degrees of conservation are indicated by colons and dots, respectively.

This study presented data on the genetic determinants of millericin B production and putative self-protection (immunity) of the producer strain. The arrangement of the millericin operon resembles that of bacteriocin systems in gram-positive bacteria. The genes required for production, secretion, and immunity are all located in the same transcription polarity. We show that unlike lysostaphin and ALE-1, millericin B appears to have a dedicated transporter (MilT) for secretion into the supernatant. This to our knowledge is the first report of a dedicated transporter for a cell wall hydrolase. Although millericin B and lysostaphin are functionally related there are several structural differences. The N-terminal tandem repeats observed in the preprolysostaphin molecule are absent in millericin B. Millericin B has a broader target specificity than lysostaphin. It has been suggested that the tandem repeats in lysostaphin allow the enzyme to interact with sites on the staphylococcal cell walls (50), and this property might contribute to its narrow spectrum of activity. Furthermore, millericin B cleaves two separate bonds within the peptide moiety of susceptible peptidoglycans, whereas lysostaphin only cleaves bonds within the peptide cross bridges.

The PCR fragments and their nucleotide sequences analyzed did not appear to contain any transcriptional stop signals. We acknowledge that while a single change in the cross bridges is quite likely necessary for resistance it may not be the only factor required for complete resistance. There may thus be other genes involved that have not yet been identified.