INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Research into the production of antibiotic-like activities by Staphylococcus aureus dates back well into the last century (8). Several groups demonstrated that bacteriocin production was especially common in S. aureus phage group II, particularly in strains of phage type 71 (5, 23). Dajani et al. (6) partially purified an inhibitory agent from tryptic soy broth cultures of S. aureus C55, a strain that they adopted as the prototype of the phage group II bacteriocin producers. The purification protocol essentially involved ammonium sulfate precipitation followed by gel filtration on Sephadex G-100. The spectrum of inhibitory activity of the agent was shown to be similar to that of bacteriocins produced by other gram-positive bacteria in that it included strains of a wide variety of other gram-positive genera. Of interest was the finding that 21 of 31 S. aureus strains but none of 18 strains of S. epidermidis appeared to be sensitive to the inhibitor (6). Mode-of-action studies indicated that cell death was due to pore formation in the cytoplasmic membrane and widespread inhibition of macromolecular biosynthesis following exposure to the partially purified material (7).

Among the most thoroughly studied staphylococcal bacteriocins in recent years are the S. epidermidis products epidermin (1), Pep5 (25), and epilancin (33). All of these are classified as lantibiotics, i.e., low-molecular-weight, posttranslationally modified peptides containing the distinctive amino acids lanthionine and/or -methyllanthionine and sometimes dehydro amino acids (29). As a consequence of these modifications, the processed, biologically active lantibiotics are relatively short, flexible, and thermally stable molecules. Several other bacteriocins isolated from S. epidermidis have subsequently been shown to be either epidermin or epidermin variants (9, 14, 26). There appears to be only a single report to date of lantibiotic production by S. aureus (31). Amino acid sequencing of this molecule, named staphylococcin Au-26, showed that there was an isoleucine residue at the N terminus followed by a sequencing-blocking residue in position 2.

In one recent study, a heat-stable, broadly active bacteriocin with a molecular mass of 6,418 Da was isolated from S. aureus KSI1829 and partially characterized (3). More recently, the same group has partially characterized another broad-spectrum S. aureus bacteriocin, namely, staphylococcin BacR1 (4, 23). BacR1 had a mass of 3,338 Da, and analysis of its amino acid composition suggested a large number of glycine residues and a relatively high proportion of hydrophobic amino acids.

Enhanced bactericidal effects due to the complementary activity of two peptides has been described for some bacteriocin complexes, including lactococcin G (20), plantaricin A (21), and plantaricin S (16). Similarly, cytolysins CylL_L and CylL_S produced by Enterococcus faecalis have also been reported to complement each other in antibacterial activity (10). This dual-peptide system is unique in that both component molecules have been identified as lantibiotics.

In the present study, we have re-examined the nature of the bactericidal activity of S. aureus C55 and demonstrated that three antibacterial peptides are produced. Two of these peptides, staphylococcins C55 and C55, were shown to be lantibiotic molecules that act synergistically to kill strains of S. aureus but not S. epidermidis.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Bacterial strains, maintenance of cultures, and detection of antibacterial activity. The parent inhibitor-producing strain is the phage group II type 71 strain S. aureus C55 described by Dajani and Wannamaker (5). Strain C55c is a derivative of C55 isolated following growth of the parent strain at 42°C in tryptic soy broth (Difco). C55c differs from C55 in the absence of a 32.0-kb plasmid and in the markedly reduced width of the inhibition zone that it produces against Micrococcus luteus T-18 in simultaneous-antagonism tests on Columbia agar base (CAB; GIBCO, Ltd., Paisley, United Kingdom). Strain T-18 is routinely used in this laboratory as a sensitive indicator of bacteriocin activity (24). S. aureus UT0007 was obtained from Pat Schlievert, University of Minnesota. All of the staphylococcal strains used to assess the activity spectrum of strain C55 were obtained from our laboratory culture collection and maintained at 4°C on CAB with 4% human blood (blood agar) when in regular use.

Purification of the inhibitory peptides. A 10-ml overnight culture of strain C55 was inoculated into 800 ml of tryptic soy broth (Difco) and then divided into two 1-liter flasks for incubation at 37°C with shaking at 160 rpm for 18 h. The cells were removed by centrifugation at 4,000 × g for 20 min, and ammonium sulfate was then slowly added to the supernatant to achieve 60% saturation. The precipitated material was collected by centrifugation and redissolved in 400 ml of PBS (pH 6.5). This preparation was applied to an XAD-2 (Serva, Heidelberg, Germany) column with a bed volume of 500 ml equilibrated in 1 mM potassium phosphate (pH 6.5). The column was washed with 3 bed volumes of 1 mM potassium phosphate (pH 6.5), followed by 4 bed volumes of 70% methanol containing 1 mM potassium phosphate (pH 6.5). These fractions were discarded. Inhibitory activity was recovered from the column by washing with 2 bed volumes of 90% methanol (adjusted to pH 2 with 11.6 M HCl) and concentrated by evaporation at 50°C under reduced pressure. Aliquots (1 ml) of this material were applied to a Brownlee C₈ reversed-phase column (Aquapore RP 300; pore size, 7 µm; 30 by 4.6 mm; Applied Biosystems, Inc.) equilibrated with 0.1% trifluoroacetic acid (TFA). Fractionation of the material was achieved by application of a stepped gradient (0 to 40% acetonitrile containing 0.085% TFA) using a Pharmacia fast protein liquid chromatography (FPLC) system at a flow rate of 1 ml/min. Each 1-ml fraction was tested for inhibitory activity against M. luteus T18. Inhibitory activity was detected in three distinct regions of the eluent fluid, and the active fractions in each region were separately pooled and named A, B, and C. Each pool was lyophilized and then dissolved in 0.1% TFA. Aliquots of each of these preparations were then loaded onto a C₁₈ reversed-phase high-pressure liquid chromatography (HPLC) column (Alltech Nucleosil C₁₈; 10 µm; 250.0 by 4.6 mm) equilibrated with 0.1% TFA and further fractionated by using a Waters/Millipore HPLC system by application of appropriate gradients of acetonitrile: for staphylococcin C55, this was 24 to 29% acetonitrile over 60 min, and for staphylococcin C55, it was 28 to 34% acetonitrile over 60 min. Inhibitory activity was detected by the CAB spot diffusion test using M. luteus T-18 as the indicator. For some assays, subinhibitory amounts of peptide pool A, B, or C were added to the individual column eluent fractions to potentially increase the sensitivity of the assay due to synergistic inhibitory activity between particular combinations of peptides. The amounts of protein in samples were determined by the modified Lowry method (22).

Electrophoresis. Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) was performed by using HPLC-purified preparations of staphylococcins C55 and C55. The preparations of staphylococcin C55 used for SDS-PAGE were obtained by lyophilization of active fractions from C₈ FPLC separations. A Mini-Protean II electrophoresis system (Bio-Rad, Richmond, Calif.) with 16.5% tricine gels (Bio-Rad) and the tricine buffer system was utilized as described by Schägger and Jagow (28). Following electrophoresis, the gels were stained with Coomassie brilliant blue R-250 (Sigma). Localization of inhibitory peptides was achieved by overlayering the gels with a thin film of CAB seeded with M. luteus T-18 and then incubating them to allow inhibitory zones to develop within the growing indicator lawn.

Antibacterial spectrum of strains C55 and C55c. The antibacterial spectrum of strains C55 and C55c was determined by using the deferred-antagonism test (15, 32). After overnight growth of S. aureus C55 as a 1-cm-wide diametric streak culture on CAB, the growth was removed and the agar surface was sterilized with chloroform vapor as described above. The strains to be tested for sensitivity to C55 products were swabbed across the line of the original streak by using overnight Todd-Hewitt broth cultures. After 18 h of incubation, any interference with the growth of the indicator in the vicinity of the original diametric growth zone was considered to be due to inhibitory agents produced by the C55 culture.

Synergy assays. HPLC-purified staphylococcin C55 and C55 preparations were assayed for inhibitory activity by the agar diffusion test. Stock 4 µM preparations of each peptide were prepared in distilled water by using lyophilized, HPLC-purified material. Assessment of the amount of protein was done by amino acid analysis. A series containing differing proportions of the two peptides was prepared in a microtiter tray, and 20-µl samples (and twofold PBS [pH 6.5] dilutions of these samples) were then spotted onto CAB plates, dried, and seeded as described above with M. luteus T-18 to detect the amount of inhibitory activity.

Plasmid DNA isolation and protoplast transformation. Plasmid DNA was purified from strains C55 and C55c by using a commercial plasmid preparation kit (QIAGEN, Hilden, Germany) as described previously (34). For large-scale plasmid extraction, a QIAGEN-tip 100 column was used. Transformation of strain C55c with plasmid DNA from strain C55 was done by using the protoplast transformation method described by Götz and Schumacher (11). Protoplasts were regenerated in modified DM 3 agar, and overlayers of soft agar containing M. luteus were used to detect transformants.

Peptide analysis. Amino acid compositions, mass spectrometry, and N-terminal amino acid sequencing were done by the Protein Microchemistry Facility, Department of Biochemistry, University of Otago. The amino acid compositions of phenylthiocarbamyl derivatives were determined by using a narrow-bore binary reversed-phase HPLC system (12). Determination of lanthionine or -methyllanthionine was done by use of the o-phthaldialdehyde method (26). Mass spectrometry was done with matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass analyzer (Finnigen Lasermat 2000; Thermo Bioanalysis), and N-terminal amino acid sequencing was done by automated Edman degradation on a 470A pulsed liquid protein sequencer (Applied Biosystems Inc.) (13).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Production and recovery of inhibitory activity. C55 cultures grown at 37°C in tryptic soy broth gave a reliable yield of inhibitory activity. Increased production over that recovered from unshaken cultures was obtained when the cultures were rotated at 160 rpm. Decreased recovery was obtained beyond 18 h of incubation, indicating some possible stationary growth phase degradation of the inhibitory activity.

Purification of antibacterial peptides. The first step toward purification of the inhibitory agents from tryptic soy broth culture supernatants involved the use of an ammonium sulfate precipitation step. Subsequent binding of the inhibitory agents to XAD-2 resins was greatly enhanced by introduction of this procedure. Elution of the inhibitory peptides bound to XAD-2 was effected by washing with 90% methanol (pH 2). The total inhibitory activity following the XAD-2 step was not significantly less than the total activity present in the original culture supernatant (Table 1).

View this table: [in this window] [in a new window]   TABLE 1.   Stepwise purification of staphylococcin C55 peptides , , and

View larger version (20K): [in this window] [in a new window]   FIG. 1.   (Panel A) C8 reversed-phase chromatography elution profile of a C55 preparation obtained by XAD-2 chromatography of culture supernatant. Protein material eluted in an acetonitrile gradient was detected by A280 measurement, and the inhibitory activity of each 1-ml fraction was tested against M. luteus T-18. Solid bars A, B, and C represent the inhibitory fractions. (Panels B and C) C18 reversed-phase HPLC analysis of the A and B inhibitory fractions obtained by FPLC (as shown in panel A). Protein material eluted in an acetonitrile gradient was detected by A214 determination, and the inhibitory activity of each fraction was tested against M. luteus T-18. In addition to major peak A, two minor peaks (A' and A") on the leading edge of peak A were found to have inhibitory activity against M. luteus T-18. Two peaks B (B and B') had inhibitory activity against M. luteus T-18 (shown by the solid bars).

Inhibitory activity against staphylococci. The range of inhibitory activity of strain C55 against staphylococci was determined by using the deferred-antagonism test. With the exception of one strain of S. carnosus, none of 52 coagulase-negative staphylococci was inhibited by strain C55 in these tests. By contrast, 120 strains of S. aureus, including several multiantibiotic-resistant and mupirocin-resistant strains, were strongly inhibited. Only two S. aureus strains were resistant to strain C55 in these tests. Both belonged to phage group II, and both appeared to produce inhibitory activities closely similar to that produced by strain C55, as judged by the identical inhibitory profiles given by all three strains in deferred-antagonism tests and by the nonsusceptibility (or immunity) that each exhibited to its own inhibitory products and to the products of the other two strains. By contrast, strain C55c (see later), an inhibitor-deficient derivative of strain C55, exhibited partial sensitivity to all three phage group II producer strains in deferred-antagonism tests.

Synergistic activity of staphylococcins C55 and C55. When HPLC-purified preparations of staphylococcins C55 and C55 (peaks A and B, respectively) were tested for inhibitory activity, this was only evident for C55. When drops of both preparations were placed adjacent to one another in a CAB assay, an enhanced zone of inhibition was produced in the region between the C55 and C55 preparations (Fig. 2). When purified peptide preparations were mixed in different concentrations, a marked increase in inhibitory activity was noticed when the peptides were present at equimolar concentrations (Table 2).

View larger version (150K): [in this window] [in a new window]   FIG. 2.   Synergistic inhibitory activity demonstrated on CAB when paper discs containing purified preparations of C55 (left) and C55 (right) were placed on a growing lawn culture of M. luteus T-18.

View this table: [in this window] [in a new window]   TABLE 2.   Inhibitory activities of mixtures of staphylococcins C55 and C55 against M. luteus T-18

Mass spectrometry. HPLC-purified staphylococcin C55 and C55 each gave clean single peaks corresponding to masses of 3,339.7 ± 0.3 and 2,993.4 ± 0.4 Da, respectively (Fig. 3 and 4).

View larger version (11K): [in this window] [in a new window]   FIG. 3.   MALDI-TOF mass spectrometry of staphylococcin C55. Glucagon (molecular mass of 3,482.8 Da) was used as the internal standard. An average molecular mass of 3,339.7 ± 0.3 Da was calculated.

View larger version (12K): [in this window] [in a new window]   FIG. 4.   MALDI-TOF mass spectrometry of staphylococcin C55. Glucagon (molecular mass of 3,482.8 Da) was used as the internal standard. An average molecular mass of 2,993.4 ± 0.4 Da was calculated.

Amino acid composition. Table 3 shows the predicted amino acid compositions of C55 and C55. Only the unmodified amino acids are listed, since the modified amino acids were not identifiable by their phenylthiocarbamyl derivatives. Both peptides were confirmed to contain lanthionine and/or -methyllanthionine by the o-phthaldialdehyde method.

View this table: [in this window] [in a new window]   TABLE 3.   Amino acid compositions of staphylococcins C55 and C55

Protein sequence analysis. Staphylococcins C55 and C55 were subjected to automated Edman degradation to determine their N-terminal amino acid sequences. The first two residues were not clearly detected in C55, and the third residue was blocked, suggesting the presence of a dehydro amino acid residue, which is a characteristic feature of lantibiotics. Digestion with chymotrypsin was not successful in giving any additional sequencing information (results not shown), but modification by using mercaptoethanol successfully unblocked the residue in position 3 and enabled further sequencing through to the 18th residue as follows: Xaa-Xaa-Dhb-Asn-Xaa-Phe-Dha-Leu-Xaa-Asp-Tyr-Trp-Gly-Asn-Lys-Gly-Asn-Trp-Xaa-Xaa-Ala-Ala. The third residue was clearly identified as Dhb, because of its characteristic chromatographic profile (27). Xaa indicates cycles in which no known amino acid derivative was detected.

Plasmid curing studies and protoplast transformation. S. aureus C55 contains two plasmids with sizes of approximately 32.0 and 2.5 kb. Incubation of tryptic soy broth cultures of strain C55 at 42°C for 18 h resulted in widespread (>90% of those tested) loss of the ability to produce large zones of inhibitory activity when individual CAB-grown colonies subcultured from these 42°C cultures were tested against M. luteus in simultaneous-antagonism stab tests (Fig. 5). Examination of the plasmid contents of representative inhibitor-deficient clones showed that they had all lost the 32.0-kb plasmid but not the 2.5-kb plasmid. Strain C55c was adopted as a representative inhibitor-deficient, 32-kb plasmid-"cured" derivative of strain C55. Strain C55c differed from the parent C55 in that it failed to inhibit the growth of strains of S. aureus in either simultaneous- or deferred-antagonism tests. Protoplast transformation of strain C55c by using purified 32-kb plasmid DNA resulted in regeneration of the ability to produce levels of staphylococcins C55 and C55 that were inhibitory to S. aureus. By contrast, all attempts to convert S. carnosus to inhibitor production by protoplast transformation met with failure.

View larger version (162K): [in this window] [in a new window]   FIG. 5.   Simultaneous-antagonism inhibitory zones produced in a lawn of M. luteus T-18 by stab cultures of S. aureus C55 (top), C55c, a "cured" derivative of strain C55 (bottom right), and strain 46 (an unrelated bacteriocin-negative isolate) (bottom left).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

This study has identified three peptides (staphylococcins C55, C55, and C55) that are responsible for the inhibitory activity of S. aureus C55. The homogeneity of the purified staphylococcin C55 and C55 preparations was established by reversed-phase HPLC, mass spectrometry, and amino acid sequencing. It was also shown that these two peptides act synergistically to inhibit the growth of strains of M. luteus and S. aureus. By contrast, there was no evidence of antibacterial interaction of staphylococcin C55 with either C55 or C55. Purified staphylococcins C55 and C55 by themselves had only relatively weak activity against the most sensitive indicator, M. luteus T-18. Combination of similar quantities of C55 and C55 increased inhibitory activity by 128-fold. Inhibitory activity against S. aureus was only detected in combinations of the two peptides.

One of the main problems in determining the amino acid sequences of lantibiotics is the frequent presence of dehydro amino acids in these molecules, leading to blocking of the Edman degradation reaction. The usual method of proteolytic digestion to obtain peptide fragments for sequencing does not always give helpful results due to the presence of lanthionine or -methyllanthionine rings, which can further alter the arrangement of the peptide or interfere with its enzymatic digestion. Combination of Edman degradation and mass spectroscopy following alkaline ethanethiol treatment was successfully applied to determine the amino acid sequences of nisin and Pep5 (18). More recently, the mutacin B-Ny 266 sequence was also resolved by use of the same treatment (19). In the present study, both staphylococcins C55 and C55 were successfully "unblocked" by use of mercaptoethanol treatment. Oligonucleotide probes based upon the sequences of staphylococcins C55 and C55 have now been used to detect the structural genes of both peptides in strain C55 and have confirmed the amino acid sequencing results (unpublished data).

The two additional well-separated peaks having inhibitory activity that eluted from the C₁₈ column in close association with the staphylococcin C55 peak probably represent variants of the predominant peptide molecule. Similarly, staphylococcin C55 preparations appear to contain another variant of the molecule (peptide C55'). The estimated masses of these additional peaks were 3,356.5 and 3,009.0 Da, respectively (data not shown in Results), and are thought to represent oxidized or incompletely dehydrated forms of staphylococcins C55 and C55 that are nevertheless still capable of interfering with the growth of M. luteus. The amount of these additional peptides recovered was not enough to enable comparison of their specific activities (arbitrary units per milligram) or any further characterization of the molecules.

The genetic determinants of previously characterized staphylococcal bacteriocins have been either plasmid or chromosome associated (2, 17, 30). In the current study, it has been shown that a 32.0-kb plasmid is required for the production of staphylococcins C55 and C55. As strain C55c displays greater sensitivity than parent strain C55 to the inhibitory products of strain C55 in deferred-antagonism tests or to an equimolar mixture of staphylococcins C55 and C55 (results not shown), it is presumed that some immunity to both of these bacteriocins is also conferred upon the cell by 32.0-kb plasmid-encoded gene products. The inability to obtain inhibitor-producing transformants of S. carnosus following protoplast transformation with 32-kb plasmid DNA from strain C55 may be due to either a lethal effect of the bacteriocins or the incompatibility of the plasmid with this S. carnosus strain. Due to our failure to identify any other marker which could be used to detect the presence of this plasmid, we were unable to establish the efficiency of the transformation process.

Crupper et al. have recently reported the purification of BacR1 from S. aureus UT0007 (4). The molecular mass of BacR1 was reported to be 3,338 Da, a size close to that of staphylococcin C55. This suggests the possibility that the same peptide has been independently isolated in the two studies. The blockage to Edman degradation encountered in the attempted sequencing of BacR1 hints at the presence of a dehydro amino acid similar to that found in staphylococcin C55. The possibility exists that the five residues obtained on sequencing BacR1 may be of a contaminating peptide species due to sequencing blockages occurring at the N terminus of the BacR1 molecule. We have cross-tested strains UT0007 (the producer of BacR1) and C55 and have found that each is immune to the inhibitory products of the other. Furthermore, an oligonucleotide probe specific for the N-terminal sequence of staphylococcin C55 was found to bind to 32.0-kb plasmid DNA present in both strains C55 and UT0007 (results not shown). The marked loss of inhibitory activity reported by Crupper et al. on reversed-phase chromatography of BacR1 may be due to the inadequate sensitivity of the Corynebacterium renale indicator strain used in that study for detection of inhibitory activity equivalent to that of staphylococcin C55. The individual components of the bacteriocin cluster reported here to be produced by S. aureus C55 may possibly be found in other staphylococcal strains. We are currently using oligonucleotide probes specific for staphylococcins C55, C55, and C55 to examine the extent of the distribution of the corresponding structural genes throughout the genus Staphylococcus.