The signal transducer (BlaRI) and the repressor (BlaI) of the Staphylococcus aureus ß-lactamase operon are inducible

Simon R. Clarkea,1 and Keith G. H. Dyke1

Microbiology Unit, Department of Biochemistry, South Parks Road, Oxford OX1 3QU, UK1

Author for correspondence: Keith Dyke. Tel: +44 1865 275293. Fax: +44 1865 275297. e-mail: kdyke{at}bioch.ox.ac.uk


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
The precise start points for transcription of the blaZ and of the blaRI/blaI genes of the Staphylococcus aureus ß-lactamase operon have been determined by primer extension analysis. Consequently the overlapping promoter sequences were deduced. Northern blots showed that the synthesis of the 2100 nt mRNA from blaRI is inducible and that a blaI probe hybridized to the same mRNA as the blaRI probe. The gene cat, encoding chloramphenicol acetyltransferase, was fused separately to the blaZ and blaRI/blaI promoters, and used to compare their strengths. The promoter for blaZ is about six times stronger than that for blaRI/blaI and the synthesis of chloramphenicol acetyltransferase from both promoters is inducible, supporting the results from the Northern blots.

Keywords: penicillin resistance, promoter strength, transcription

Abbreviations: CAT, chloramphenicol acetyltransferase; CBAP, 2-(2'-carboxyphenyl)-benzoyl-6-aminopenicillanic acid; MRSA, methicillin-resistant Staphylococcus aureus

a Present address: Department of Molecular Biology and Biotechnology, Firth Court, University of Sheffield, Western Bank, Sheffield S10 2TN, UK.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
The human pathogen Staphylococcus aureus is a major cause of nosocomial infections and treatment often fails because the causative bacteria are resistant to antibiotics, especially ß-lactams. Such resistance is either because the organism produces ß-lactamase that hydrolyses the ß-lactam or because a penicillin-binding protein known as PBP2a, which has reduced affinity for ß-lactams, is produced. In the latter case the organisms are described as methicillin-resistant or MRSA.

The region responsible for synthesis of ß-lactamase has been sequenced (Rowland & Dyke, 1989 , 1990 ) and shown to consist of blaZ (the gene coding the ß-lactamase), blaRI (the gene encoding a putative transmembrane signal transducer) and blaI (the gene encoding the repressor). The genes blaRI/blaI are believed to be transcribed as a single mRNA in the opposite direction to the transcription of blaZ (Gregory et al., 1997 ). This is very similar to the arrangement found in the pen region of Bacillus licheniformis (Kobayashi et al., 1987 ) except that the order of transcription of the homologues of blaRI (penJ) and blaI (penI) is reversed, with penI nearest to the promoter in B. licheniformis. This difference in the order of the genes is perhaps the reason for the difference in the kinetics of induction. In S. aureus the ß-lactamase is fully induced in just a few minutes whereas in B. licheniformis full induction does not occur until about 90 min (Collins, 1979 ). Transcription of penI and penJ has been studied, and in the first 30 min of induction 2·3 kb transcripts encoding penI and penJ are produced but these are then replaced by shorter transcripts of 0·75 and 0·95 kb that encode penI only (Salerno & Lampen, 1988 ). They suggested that PenJ was required early in the induction process. It is not known if different sized mRNAs encoding BlaI are produced on induction of ß-lactamase in S. aureus. The induction of ß-lactamases in bacteria has been reviewed by Bennett & Chopra (1993) .

The mechanism of induction of the ß-lactamase in S. aureus is beginning to be understood and is based on proteolytic cleavage of the repressor, BlaI, that is dependent on the presence of BlaRI (Gregory et al., 1997 ; Lewis et al., 1999 ). The hypothesis is that a ß-lactam binds to the extracellular part of BlaRI and causes a conformational change that activates the intracellular proteolytic activity of BlaRI. The repressor, BlaI, is cleaved by the BlaRI protease so that it no longer binds to the bla operator, thus allowing synthesis of blaZ mRNA and hence ß-lactamase. It is not obvious that there has to be continuous synthesis of BlaI and BlaRI during induction and there is evidence from Western blotting that BlaI is not significantly induced (Lewis et al., 1999 ).

The mechanism of induction may also apply to the synthesis of PBP2a from mecA of MRSA. In some MRSA strains the synthesis of PBP2a is inducible and controlled by two genes: mecRI and mecI, which are homologous to blaRI and blaI (Hiramatsu et al., 1992 ). Indeed it has been shown that BlaI can repress the synthesis of MecA (Hackbarth & Chambers, 1993 ) and that MecI can repress synthesis of ß-lactamase (Lewis & Dyke, 2000 ). Thus any unravelling of the mechanism of induction of ß-lactamase will probably also apply to the mechanism in at least some MRSA strains. In other MRSA strains the blaRI or blaI genes carry inactivating mutations but despite this, in the presence of the bla operon, the synthesis from mecA is inducible. There is an urgent need for new therapies to control infections by MRSA and it is possible that better knowledge of the mecA induction system could eventually lead to the design of alternative therapies.

In this paper the results of experiments designed to discover, in particular, whether or not the synthesis of BlaRI and BlaI is inducible and to find out if they share the same promoter are reported. The techniques used include Northern blotting and the use of the cat gene as a reporter of synthesis from the promoters. The promoters for both blaZ and blaRI/blaI were found to be inducible and evidence is presented to show that the blaZ promoter is about six times stronger than that for blaRI/blaZ.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Bacteria, plasmids and growth medium.
Staphylococcus aureus RN4220 (Kreiswirth et al., 1983 ), which is a modified NCTC8325 strain that accepts Escherichia coli DNA, was used as a recipient in transformation by the method of Gotz et al. (1981) . E. coli JM109 (Yanisch-Perron et al., 1985 ) was used as recipient in transformation by the method of Hanahan (1983) . The plasmids used in this study are listed in Table 1. S. aureus was grown at 30 °C in CY medium (Novick, 1963 ) with aeration or on CY agar plates. E. coli was grown in Luria–Bertani (LB) medium at 37 °C.


View this table:
[in this window]
[in a new window]
 
Table 1. S. aureus strains and plasmids used in this study

 
Oligonucleotides.
These are listed in Table 2.


View this table:
[in this window]
[in a new window]
 
Table 2. Oligonucleotides used in this study

 
PCR.
PCRs were carried out with Taq polymerase and with 25 cycles of 1 min at 95 °C, 1 min at 50 °C and 90 s at 70 °C followed finally by 8 min at 70 °C.

Preparation of RNA.
S. aureus was grown to an OD680 of 0·6 and, if required, induced for various times by continuing growth in the presence of 5 µg 2-(2'-carboxyphenyl)-benzoyl-6-aminopenicillanic acid (CBAP) ml-1. RNA was extracted from the harvested bacteria by the method of Hart et al. (1993) except that the protoplasts were lysed with 5 µl SDS (2%, w/v) and 50 µl proteinase K (5 mg ml-1). The pellet of RNA was resuspended in 100 µl nuclease-free water. All plastic and glassware used for the preparation and storage of RNA was treated with RNase ZAP (Sigma), rinsed with 0·1% (v/v) diethyl pyrocarbonate and autoclaved at 121 °C, 103·5 kPa for 20 min.

Labelling of oligonucleotides and probes.
The methods described by Sambrook et al. (1989) were used for end labelling oligonucleotides with 32P by the T4 polynucleotide kinase forward reaction and for analysis of RNA by primer extension. The Random Primers DNA Labelling System (Life Technologies) was used to label the probes for Northern blotting.

Northern blotting.
RNA samples were heated at 65 °C for 5 min, cooled on ice and then loaded onto a 1·5% (w/v) agarose-MOPS [20 mM MOPS, 5 mM sodium acetate, 0·5 mM EDTA] denaturing gel containing 6·5% (w/v) formaldehyde, and electrophoresed at 80 V. After electrophoresis, the RNA was transferred to a positively charged nylon membrane by capillary blotting in 10x SSC. The RNA was fixed to the membrane by UV crosslinking. The membrane was pre-hybridized at 42 °C for 2 h in 20 ml pre-warmed hybridization buffer [50% (w/v) deionized formamide, 5x SSC, 2% (w/v) blocking agent (Boehringer Mannheim), 0·1% (w/v) N-laurylsarcosine, 0·02% (w/v) SDS, 0·1 mg boiled calf thymus DNA ml-1].

Hybridization with heat-denatured radiolabelled probe was for 18 h in 2 ml hybridization buffer. The membrane was washed twice at room temperature in 100 ml membrane wash (2x SSC; 0·1%, w/v, SDS) for 5 and 15 min consecutively. Bands were visualized by autoradiography.

DNA sequencing.
Sequenase version 2.0 DNA sequencing kit (USB) was used for manual DNA sequencing. The M13Forward primer (Table 2) was used to produce a sequence to locate the transcription start sites for blaZ and blaRI/blaI.

Construction of cat gene reporter plasmids.
Plasmids in which the cat gene was driven by either the blaZ or blaRI/blaI promoter were constructed to allow comparison of both the strengths of the promoters and their responses to induction. Plasmid pOX690 has a cat gene driven by the blaRI/blaI promoter (Gregory et al., 1997 ) but does not have the blaRI and blaI genes that are essential to study the induction from the blaRI/blaI promoter. To provide BlaRI and BlaI, pOX491 (Fig. 1) was digested with AvaII/AccI, the ends filled in with the Klenow polymerase reaction, religated and used to transform E. coli JM109 with selection for ampicillin resistance. The resulting plasmid (pSRC900) was then digested with PstI and the 3·6 kb piece containing the three bla genes was ligated into PstI-digested pOX690 and used to transform E. coli JM109 with selection for ApR to give pSRC910. A similar cloning strategy, but starting with pOX617 (Gregory et al., 1997 ) in which the bla operator had been cloned into the pUC18 SmaI site, was followed to produce a plasmid which contains the three bla genes and in which the cat gene is driven by the blaZ promoter. First, both pOX617 and pRB394 (Bruckner, 1992 ) were cut with EcoRI/PstI, ligated and used to transform E. coli to chloramphenicol resistance. A clone that had the operator region ligated into pRB394 so that the blaZ promoter drives the cat gene was chosen and the plasmid called pSRC920. DNA of pSRC920 was cut with PstI and ligated to the 3·6 kb PstI fragment from pSRC900. The resulting plasmid is pSRC921 and contains a complete bla operon in addition to the blaZ promoter that drives the cat gene. Restriction enzyme maps of pSRC910 and pSRC921 are shown in Fig. 2. So that the plasmids can replicate in S. aureus, both pSRC910 and pSRC921 were digested with BglII and separately ligated to BamHI-digested pT181MCS (Augustin et al., 1992 ). The ligation mixtures were used to transform E. coli JM109 with selection for ampicillin resistance and the resulting plasmids were then used to transform S. aureus RN4220 to tetracycline resistance. A plasmid from each transformation which had pT181MCS inserted in the same orientation was chosen and designated pSRC910::pT181MCS and pSRC921::pT181MCS respectively.



View larger version (13K):
[in this window]
[in a new window]
 
Fig. 1. Restriction enzyme map of pOX491. Only a selection of restriction enzyme sites are shown. CmR indicates the gene encoding resistance to chloramphenicol, EmR indicates the gene encoding resistance to erythromycin. Rep is the replication protein of pE194. The blaZ, blaRI and blaI genes are marked and the arrows represent the direction of transcription. binL and P480 are genes of Tn552.

 


View larger version (17K):
[in this window]
[in a new window]
 
Fig. 2. Restriction enzyme maps of the plasmids pSRC921 and pSRC910. PZ and PRI are the promoters for blaZ and blaRI/blaI respectively. ApR, ampicillin resistance; KmR, kanamycin resistance; cat, promoterless gene for chloramphenicol acetyltransferase.

 
Chloramphenicol acetyltransferase (CAT) assays.
Reaction conditions were adapted from Tomizawa (1985) . A mid-exponential phase culture (0·5 ml) was harvested and the bacteria resuspended in 0·5 ml 0·1 M Tris/HCl (pH 8·0). Twenty microlitres of lysis buffer (0·1 M EDTA, 0·1 M DTT, 50 mM Tris/HCl, pH 8·0) and a small drop of toluene were added and the mixture vortexed for 30 s. The sample was incubated at 30 °C for 30 min to evaporate the toluene. Ten microlitres of the resulting mix was incubated with 100 µl 100 µM chloramphenicol and 2 µl [1-14C]acetyl-CoA (2·22 GBq mmol-1) at 30 °C for 10 min. The reaction was terminated by the addition of 0·2 ml toluene, into which the radiolabelled [14C]acetylchloramphenicol partitions. After vigorous mixing, the phases were separated by centrifugation and 100 µl of the organic phase was counted in a Beckman LS 5000TD scintillation counter.


   RESULTS AND DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Identification of transcriptional start sites
Oligonucleotides 5448SRC and 5270SRC were used as primers in primer extension analysis to map the positions of the start sites for the blaZ and blaRI/blaI mRNAs respectively. The sources of RNA were four strains of S. aureus: RN4220 as a negative control;NCTC9789(pI9789) which has a plasmid carrying the bla operon (Asheshov, 1966 ); PS80d which is a ß-lactamase-negative strain in which a copy of Tn552 is integrated into the chromosome (Sohail & Dyke, 1996 ); and 3796(pI3796) in which a base is deleted so altering the reading frame for BlaZ (East & Dyke, 1989 ).

The results for blaZ (Fig. 3) show that the start site for mRNA synthesis is the same for PS80d (lane 1), NCTC9789(pI9789) (lane 2) and 3796(pI3796) (lane 3), and is at a thymidine as shown in Fig. 4. This thymidine is one base away from the site identified for blaZ mRNA by an in vitro transcription system using E. coli RNA polymerase (McLaughlin et al., 1981 ). The -10 sequence for the blaZ promoter is likely to be TATTAT which is precisely in the expected position and the -35 sequence is probably TTGACA which is separated from the -10 sequence by 18 bp. The untranslated part of this mRNA contains the Z dyad which is one of the sites on the DNA to which BlaI is known to bind (Gregory et al., 1997 ). The second dyad (RI dyad) to which BlaI binds lies between the -35 sequence and the -10 sequence, and partly overlaps the latter (Fig. 4). These data fit with the hypothesis that binding of the repressor BlaI at this dyad will prevent the binding of RNA polymerase and with the experimental observation that deletion of the Z dyad increases transcription from the blaZ promoter (S. R. Clarke, unpublished results). PS80d does not produce ß-lactamase but it must possess at least the first section of the blaZ gene including the promoter. It therefore seems likely that there is a mutation that prevents the synthesis of active ß-lactamase. The data for 3796(pI3796) are the same as for NCTC9789(pI9789) and the pI3796 DNA used here was found not to have the deletion at the beginning of the blaZ gene. The reason for the discrepancy with the results of East & Dyke (1989) is unknown since the earlier reported sequence was not found with the cultures now available.



View larger version (93K):
[in this window]
[in a new window]
 
Fig. 3. Autoradiograph of primer extension of blaZ transcripts from S. aureus strains. Lane 1, PS80d; lane 2, NCTC9789(pI9789); lane 3, 3796(pI3796). The sequence of pUC18 directed by the M13Forward primer is in lanes A, C, G and T. The position on the pUC18 sequence corresponding to the position of the start site for the blaZ transcript is marked with an asterisk (*).

 


View larger version (9K):
[in this window]
[in a new window]
 
Fig. 4. The sequence of the bla operon between the start sites for translation of BlaZ and BlaRI. The transcription start sites are labelled +1 mRNA. The proposed -35 and -10 promoter sites are marked as are the ribosome-binding sites (RBS). The RI dyad and Z dyad sequences are in bold and their inverted repeats are indicated by arrows.

 
The primer extension results (not shown) for the blaRI/blaI mRNA place the start site at the position indicated in Fig. 4 for both NCTC9789(pI9789) and PS80d. The -10 sequence for the promoter for this mRNA is probably TATTAT which is ten bases from the start site and 17 bp from the most probable -35 sequence of TTGTAA (Fig. 4). Thus the -10 sequence may be the same for both the blaZ and blaRI/blaI promoters but the -35 sequences differ by 2 nt. Since the -35 sequence and the RI dyad overlap, it is likely that there will be competition for binding between the repressor and RNA polymerase and so it was decided to investigate whether the synthesis of blaRI/blaI specific mRNA was inducible and to discover whether there were other mRNAs encoding BlaI.

Identification of mRNA for BlaRI and BlaI by Northern blotting
The probe for blaRI mRNA was made by PCR with the primers 5106SRC (5106–5087 on the Tn552 sequence) and 3720SRC (3704–3720 on the Tn552 sequence), and with pOX491 as the template, which results in a 1403 bp probe entirely from within the blaRI sequence. The RNA was prepared from S. aureus RN4220(pOX491) which contains an intact bla operon (Fig. 1) and from S. aureus RN4220(pOX492) which was derived from pOX491 by cutting with SnaBI, ligating and then transforming into E. coli JM109 with selection for chloramphenicol resistance. The plasmid was then transformed into S. aureus RN4220 with selection for erythromycin resistance. pOX492 lacks the end of the blaRI gene and the part of Tn552 that encodes blaI, binL and (P480). The results of the Northern blot (Fig. 5) show that only a small amount of message for BlaRI is produced by RN4220(pOX491) unless the bacteria are induced and that the message is about 2100 nt long. The mRNA from pOX492 is about 1700 nt, which is about the predicted size provided that there is a suitable termination signal shortly after the SnaBI site in pOX492. Since blaI has been deleted it is to be expected that similar amounts of mRNA will be produced with and without inducer for pOX492 and this is seen (Fig. 5). Since no other bands were seen, the Northern blot provides evidence that a single mRNA is produced from the blaRI/blaI promoter and that it is inducible.



View larger version (40K):
[in this window]
[in a new window]
 
Fig. 5. Autoradiograph of a Northern blot of total RNA from strains of S. aureus. The RNAs in the induced lanes were prepared after the cultures had been induced with CBAP for 20 min. The positions of the 2·1 kb and 1·7 kb bands are indicated. The 1403 bp blaRI probe was produced from the primers 3720SRC and 5106SRC by PCR.

 
Several probes for blaI were made by PCR including one of 442 bp using oligonucleotides 3546SRC and 3105SRC (Table 2). This probe stretches from the beginning of blaI to some 60 bases after the stop codon. All the blaI probes bound to a 2100 nt RNA but did not reveal any other blaI-specific message, even after prolonged induction (data not shown). Thus the system in S. aureus differs from that found with B. licheniformis. where several species of penI mRNA were found (Salerno & Lampen, 1988 ).

Reporter gene assays
S. aureus RN4220(pSRC910::pT181MCS) and RN4220(pSRC921::pT181MCS) were grown at 30 °C for 18 h in CY medium containing 5 µg tetracycline ml-1, diluted 1:100 in the same medium and growth continued at 30 °C to an OD680 of 0·5. The gratuitous inducer CBAP (Leggate & Holms, 1968 ) was added to a final concentration of 5 µg ml-1 and samples taken every 2 min for the assay of CAT activity. The results for the expression of CAT in the absence of inducer (results not shown) are the same as those for the zero time points for addition of inducer (Table 3) so that it is concluded that, in the absence of inducer, the blaZ promoter is about six times more active than the blaRI/blaI promoter. There is rapid induction of CAT in S. aureus from both promoters (Table 3) and it seems that a balance between synthesis and degradation of CAT is attained after about 8 min induction. The reporter system used may not accurately mimic the wild-type position; for example, there are two copies of the operator region in the bacterium but only one set of active blaRI/blaI genes. However, it provides excellent evidence that synthesis from the blaRI/blaI promoter is inducible and supports the Northern blot results. A further problem could be associated with variation of copy numbers of the plasmids. In both pSRC910 and pSRC921 the copy number is controlled by the pT181MCS plasmid and the plasmids are of similar size. Care was taken to harvest the bacteria at the same density (OD680 0·6) to avoid differences in copy number at different growth phases. Thus comparison of the strengths of the two promoters is taken as representative of the state in the wild-type. However, the activity of the blaZ promoter in the absence of inducer is somewhat higher than was predicted.


View this table:
[in this window]
[in a new window]
 
Table 3. Time course of the induction of CAT from the wild-type blaZ and blaRI/blaI promoters in S. aureus RN4220(pSRC910::pT181MCS) and RN4220(pSRC921::pT181MCS)

 
The model for regulation of ß-lactamase synthesis
The results reported allow more precision to be introduced into the model (Dyke & Gregory, 1997 ) for induction of ß-lactamase in this organism. A ß-lactam binds to the extracellular part of BlaRI and causes a conformational change that activates the intracellular proteolytic activity of BlaRI which cleaves not only BlaI but also itself. This auto-proteolytic activity could explain why it is necessary to induce synthesis of BlaRI since newly synthesized molecules of BlaRI will continue to respond to extracellular ß-lactams whereas cleaved ones may not. The auto-proteolytic activity would also provide an explanation for the apparently discrepant results of Hackbarth et al. (1994) who found BlaRI to be 69 kD while Bruns et al. (1997) reported it to be 35 kD. The latter could be a derived polypeptide from BlaRI. The proteolytic activity of the activated BlaRI will cleave BlaI and so inactivate it, and permit transcription from both the blaZ and blaRI/blaI promoters. Consistent with the -10 and -35 sequences reported in this paper and with the CAT assay results, the RNA polymerase will have a higher affinity for the blaZ promoter and so will interfere with the binding of RNA polymerase to the blaRI/blaI promoter. Induction of BlaI can be explained since the concentration of active BlaI will be maintained at a low level but there will be continuous proteolytic cleavage of it until the extracellular concentration of ß-lactam is reduced. With the cessation of proteolytic cleavage of BlaI by BlaRI the concentration of BlaI will rapidly increase and so repression will be re-established with the minimum expenditure of energy in synthesizing unnecessary ß-lactamase.

It is probable that the results that apply to the regulation of blaZ will also apply to the regulation of mecA in cases where mecRI and mecI are not intact. Because of the extensive homology between the proteins it also seems likely that an analogous system will apply when the synthesis of mecA is regulated by MecRI and MecI. In the latter case, chemicals that interfered with the binding of the inducer to MecRI would inhibit the synthesis of MecA and hence inhibit the expression of methicillin resistance by the bacterium.


   ACKNOWLEDGEMENTS
 
We thank Mr Stephen Curnock for technical assistance. We are grateful to the BBSRC for an award to S.R.C., and to both the Edward Penley Abraham Fund and the British Society for Antimicrobial Chemotherapy for support.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Asheshov, E. H. (1966). Chromosomal location of the genetic elements controlling penicillase production in a strain of Staphylococcus aureus. Nature 210, 804-806.[Medline]

Augustin, J., Rosenstein, R., Weiland, B., Schneider, U., Schnell, N., Engelke, G., Entian, K. & Gotz, F. (1992). Genetic analysis of epidermin biosynthetic genes and epidermin-negative mutants. Eur J Biochem 204, 1149-1154.[Abstract]

Bennett, P. M. & Chopra, I. (1993). Molecular basis of ß-lactamase induction in bacteria. Antimicrob Agents Chemother 37, 153-158.[Medline]

Bruckner, R. (1992). A series of shuttle vectors for Bacillus subtilis and Escherichia coli. Gene 122, 187-192.[Medline]

Bruns, O., Bruns, W. & Pulverer, G. (1997). Regulation of ß-lactamase synthesis as a novel site of action for suppression of methicillin resistance in Staphylococcus aureus. Zentbl Bakteriol 285, 413-430.

Collins, J. F. (1979). ß-Lactamases of Staphylococcus aureus. In ß-Lactamases , pp. 351-368. Edited by J. M. Hamilton-Miller & J. T. Smith. London: Academic Press.

Dyke, K. G. H. & Gregory, P. D. (1997). Resistance to ß-lactam antibiotics: resistance mediated by ß-lactamase. In The Staphylococci in Human Disease , pp. 136-157. Edited by K. B. Crossley & G. L. Archer. Edinburgh: Churchill Livingstone.

East, A. K. & Dyke, K. G. H. (1989). Cloning and sequence determination of six Staphylococcus aureus ß-lactamases and their expression in Escherichia coli and S. aureus. J Gen Microbiol 135, 1001-1015.[Medline]

Gotz, F., Ahrne, S. & Lindberg, M. (1981). Plasmid transfer and genetic recombination by protoplast fusion in staphylococci. J Bacteriol 145, 74-81.[Medline]

Gregory, P. D., Lewis, R. A., Curnock, S. P. & Dyke, K. G. H. (1997). Studies of the repressor (BlaI) of ß-lactamase synthesis in Staphylococcus aureus. Mol Microbiol 24, 1025-1037.[Medline]

Hackbarth, C. J. & Chambers, H. F. (1993). blaI and blaRI regulate ß-lactamase and PBP2' production in methicillin resistant Staphylococcus aureus. Antimicrob Agents Chemother 37, 1144-1149.[Abstract]

Hackbarth, C. J., Miick, T. C. & Chambers, H. F. (1994). Altered production of penicillin-binding protein 2a can affect phenotypic expression of methicillin resistance in Staphylococcus aureus. Antimicrob Agents Chemother 38, 2568-2571.[Abstract]

Hanahan, D. (1983). Studies on the transformation of Escherichia coli with plasmids. J Mol Biol 166, 557-580.[Medline]

Hart, M. E., Smeltzer, M. S. & Iandolo, J. J. (1993). The extracellular protein regulator (xpr) affects exoprotein and agr mRNA levels in Staphylococcus aureus. J Bacteriol 175, 7875-7879.[Abstract]

Hiramatsu, K., Asada, K., Suzuki, E., Okonogi, K. & Yokota, T. (1992). Molecular cloning and nucleotide sequence determination of the regulator region of mecA gene in methicillin-resistant Staphylococcus aureus (MRSA). FEBS Lett 298, 133-136.[Medline]

Horinouchi, S. & Weisblum, B. (1982). Sequence of a small staphylococcal plasmid, pE194. J Bacteriol 150, 804-814.[Medline]

Kobayashi, T., Zhu, Y. F., Nicholls, N. & Lampen, J. O. (1987). A second regulatory gene, blaRI, encoding a potential penicillin-binding protein required for induction of ß-lactamase in Bacillus licheniformis. J Bacteriol 169, 3873-3878.[Medline]

Kreiswirth, B. H., Lofdahl, S., Betley, M. J., O’Reilly, M., Schleivert, P. M., Bergdoll, M. S. & Novick, R. P. (1983). The toxic shock syndrome exotoxin is not detectably transmitted by a prophage. Nature 305, 709-712.[Medline]

Leggate, J. & Holms, W. H. (1968). Gratuitous synthesis of ß-lactamase in Staphylococcus aureus. J Bacteriol 96, 2110-2117.[Medline]

Lewis, R. A. & Dyke, K. G. H. (2000). MecI represses synthesis from the ß-lactamase operon of Staphylococcus aureus. J Antimicrob Chemother 45, 139-144.[Abstract/Free Full Text]

Lewis, R. A., Curnock, S. C. & Dyke, K. G. H. (1999). Proteolytic cleavage of the repressor (BlaI) of ß-lactamase synthesis in Staphylococcus aureus. FEMS Microbiol Lett 178, 271-275.[Medline]

McLaughlin, J. R., Murray, C. L. & Rabinowitz, J. C. (1981). Unique features in the ribosome binding site sequence of the gram-positive Staphylococcus aureus ß-lactamase gene. J Biol Chem 56, 11283-11291.

Novick, R. P. (1963). Analysis by transduction of mutants affecting penicillinase formation in Staphylococcus aureus. J Gen Microbiol 33, 121-136.

Rowland, S.-J. & Dyke, K. G. H. (1989). Characterization of the staphylococcal ß-lactamase transposon Tn552. EMBO J 8, 2761-2773.[Abstract]

Rowland, S.-J. & Dyke, K. G. (1990). Tn552, a novel transposable element from Staphylococcus aureus. Mol Microbiol 4, 961-975.[Medline]

Salerno, A. J. & Lampen, J. O. (1988). Differential transcription of the bla regulatory region during induction of ß-lactamase in Bacillus licheniformis. FEBS Lett 227, 61-65.[Medline]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sohail, M. & Dyke, K. G. H. (1996). Suppression of the thermosensitive replication phenotype of the derivative plasmid of pI9789::Tn552 in Staphylococcus aureus may involve integration of the plasmid into the host chromosome. FEMS Microbiol Lett 136, 129-136.[Medline]

Tomizawa, J.-I. (1985). Control of ColE1 plasmid replication: initial interaction of RNA I and the primer transcript is reversible. Cell 40, 527-535.[Medline]

Yanisch-Perron, C., Vieira, J. & Messing, J. (1985). Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33, 103-119.[Medline]

Received 30 August 2000; revised 17 October 2000; accepted 5 December 2000.