Involved in Methicillin Resistance in Staphylococcus aureus

ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, July 1992, P. 1367-1373 0066-4804/92/071367-07$02.00/0

Vol. 36, No. 7

Copyright X 1992, American Society for Microbiology

Mapping and Characterization of Multiple Chromosomal Factors Involved in Methicillin Resistance in Staphylococcus aureus BRIGITTE BERGER-BACHI,* ANNI STRASSLE, JOHN E. GUSTAFSON, AND FRITZ H. KAYSER Institute of Medical Microbiology, University of Zurich, Gloriastrasse 32, CH8028 Zurich, Switzerland Received 4 October 1991/Accepted 7 April 1992

Chromosomal factors, termedfem or aux factors, are needed for the expression of methicillin resistance in methicillin-resistant (Mc') Staphylococcus aureus; also needed is the mec-encoded low-affinity penicillin-binding protein PBP 2'. These factors make up part of the normal set of genes present in susceptible and resistant strains of S. aureus and can be identified by TnSSl-mediated insertional inactivation of the methicillin resistance. In this study, we characterized different TnS51 inserts and mapped them into four distinct loci on the SmaI chromosomal map of S. aureus NCTC 8325, thereby identifying two new loci which code for fem factors. The largest fragment, SmaI-A, carries three loci, two coding for both closely linked factorsfemA and femB and a novel third locus (femC) that is not linked to the other two. An additional, fourth, locus,femD, was identified in fragment SmaI-I. femi4 and femB inactivation reduced overall methicillin resistance, whereby femB had less of an influence on the resistance level. femC and femD inactivation reduced mainly the basal resistance level in heterogeneously Mcr strains and had less of an impact on the subpopulation with high-level resistance. Inactivation of either of these factors was shown to have no influence on the production of PBP 2', the main factor mediating methicillin resistance. In addition, no changes were observed in the banding patterns of the major autolysins in whole-cell extracts of thefem mutants, suggesting that the reduced cell wall turnover and autolysis observed in some of the insertionally inactivated strains were due to changes either of the substrate or in the autolysin control.

Methicillin resistance in staphylococci is mediated by the methicillin resistance determinant mec, which carries the structural gene mecA that codes for the low-affinity penicillin-binding protein (PBP) PBP 2' or PBP 2a (14, 27, 31). This novel PBP is thought to allow for peptidoglycan biosynthesis at concentrations of ,-lactams which inactivate the preexisting PBPs 1, 2, 3, and 4 (10, 12, 21, 25) in methicillinresistant (Mcr) strains. A typical trait of mec-mediated methicillin resistance is the heterogeneous expression of resistance within a population of a Mcr strain. The majority of cells show only a low level of resistance, whereas minor subpopulations demonstrate resistance to higher concentrations of methicillin (for a review, see reference 7). The proportion of cells that express higher resistance levels is strain dependent and can vary from 10-2 to 10-8 within a single population (30). Some strains, termed homogeneously Mcr strains, contain only high-level-resistant cells. The mechanisms and genes which govern the switch from lowlevel resistance to the various higher resistance levels are as yet unknown. The level of resistance in Mcr strains does not correlate with the quantity of PBP 2' present (4, 8, 17, 18). It is known that additional chromosomally located genes that are not linked to mec and that have been identified by TnS51 insertional mutagenesis are essential for the expression of methicillin resistance (2, 4, 17). These genes, termed fem (factors essential for the expression of methicillin resistance) (3) or aux (auxiliary) (29) factors, are found in both Mcr and susceptible strains. One of these factors, femA, has been sequenced, and upon TnS51-mediated inactivation it causes a decrease in the peptidoglycan-associated glycine content (3, 19). Downstream and adjacent to femA lies another

*

factor, femB, which also affects methicillin resistance levels, although to a lesser extent than does femA (3). It has been shown that a number of Tn551 insertions in the chromosome of the homogeneously Mcr strain Col, reduce methicillin resistance (17), cell wall turnover, and autolysis (9), similar to femA inactivation (3, 19). The exact function and location of all these Tn551 insertions that affect methicillin resistance have yet to be determined. Using pulsedfield gel electrophoresis of SmaI-digested chromosomal DNA and Southern blotting, we were able to distinguish four different chromosomal loci that are not linked to the mec gene but that affect methicillin resistance. We also demonstrate that the inactivation of these factors, which affect cell wall metabolism, has no effect on the production of PBPs or major autolysins. This result suggests that these factors are involved with cell wall synthesis in some unknown manner. MATERIALS AND METHODS Bacterial strains, plasmids, and growth conditions. The bacterial strains and plasmids used in this study are listed in Table 1. The TnS51 insertions that lower methicillin resistance and that were originally present in various strains of Col that were kindly provided by John Kornblum (17) were transduced with phage 80a to Mcr strains BB270 and BB271. This was done in order to map them on the chromosomal SmaI restriction map of NCTC 8325 (23). Transductants were selected on 20 ,ug of erythromycin per ml and were screened for their inability to grow on 5 ,ug of methicillin per ml. Shuttle vector pHV33, obtained from N. El Solh, and shuttle vector pGC2, a generous gift from P. Matthews, were used to clone the femA and femB regions of Staphylococcus aureus as described earlier (3). When it was needed, chloramphenicol (20 ,ug/ml) was added to maintain recombinant plasmids. Bacteria were grown in LB medium (10 g of

Corresponding author. 1367

1368

ANTIMICROB. AGENTS CHEMOTHER.

BERGER-BACHI ET AL. TABLE 1. S. aureus strains and plasmids used in this study

Origin or reference

Relevant genetic marker

Strain

Strain Col BB403 BB671

Col, mec Col, mec fQ2003(femA::Tn551) Col, mec fQ2003(femA::Tn551), pBBB31

BB672

Col, mec fI2003(femA::TnS51), pBBB64

BB255 BB270 BB271

NCTC 8325 NCTC 8325, A(30kb::SmaI-F) mec NCTC 8325, mec

BB308 BB769

NCTC 8325, mec fI2003(femA::Tn551) NCTC 8325, A(30kb::SmaI-F) mec fI2003(femA::TnS51)

BB597 BB592 BB593 BB589

NCTC 8325, NCTC 8325, NCTC 8325, NCTC 8325,

BB591

NCTC 8325, A(30kb::SmaI-F) mec Q12F(femD::TnSSJ)

BB586 BB664

NCTC 8325, mec fQ2003(femA::TnS51), pBBB31 NCTC 8325, mec Q2003(femA::Tn551), pBBB64

Plasmid pBBB31

3; Spontaneous deletion in SmaI-F This study, essentially isogenic to BB270, without deletion in SmaI-F 3 This study, by transduction of fQ2003 from BB308 with phage 80a in BB270 3 3 3 This study, by transduction of insert fQ2005 from Col 1H (17) with phage 80ot into BB270 This study, by transduction of insert Q12F from Col 12F (17) with phage 80ax into BB270 3 This study, by transformation of BB308 with plasmid pBBB64

A(30kb::SmaI-F) mec flIII-8(femA::Tn551) A(30kb::SmaI-F) mec fII-1(femB::Tn5Sl) A(30kb::SmaI-F) mec fIII-2(femB::Tn551) A(30kb::SmaI-F) mec fQ2005(femC::TnS51)

Shuttle vector pHV33, 2.2-kb EcoRV fragment (femA+) inserta Shuttle vector pGC2, 5-kb PstI fragment (femA+, femB+) inserta

pBBB64 a

17 3 This study, by transformation of BB403 with pBBB31 This study, by transformation of BB403 with pBBB64

3

This study

See Fig. 2.

codon 95 and going downstream of the agrB gene (kindly provided by S. Arvidson), a 2-kb EcoRI-HindIII fragment covering mecA (provided by C. Ryffel), a 1.8-kb AvaI fragment covering ennB of Tn5Sl, and a 10.5-kb PstI fragment covering the femA and femB regions (3) shown in Fig. 1. PBP assays. PBP assays were performed essentially as described by Berger-Bachi et al. (5). Membranes isolated from exponentially growing cells were labeled for 10 min at 30°C with 10 ,ug of [3H]benzylpenicillin (57.83 mCi/mg; Merck Sharp & Dohme, Rahway N.Y.) per ml. Competition assays to saturate all PBPs except PBP 2' were carried out with membranes that were preincubated in the presence of

tryptone, 5 g of yeast extract, 5 g of NaCl) at 370C with

shaking. DNA manipulations. All DNA manipulations, gel electrophoresis, blotting of DNA, and hybridization procedures were performed essentially as described by Maniatis et al. (20). Restriction enzymes were purchased from Boehringer Mannheim and were used according to the recommendations of the manufacturer. Pulsed-field gel electrophoresis of SmaI chromosomal digests was performed essentially as described by Goering and Duensing (11) and was carried out by using a CHEF-DR II electrophoresis cell (Bio-Rad, Richmond, Calif.). The following DNA probes were used: a 4-kb BglII fragment of +11 (1), a 7.0-kb PstI fragment starting from

0

(3 I

I

co

< 0,)

0

0

8'. I

m

Ij _i w

0

0

w

w

0

w

a.

I I L___L___l__ I1 ......... bE EEEh- mummmmmlu.u l..uuuE E1

0

NC femA _ .

CCc

I

1 kb

feIB pBBB31

pBBB64 FIG. 1. Restriction map of thefemA-femB region. The open reading framesfemA andfemB are shown. The arrow points in the direction of transcription of the polycistronic mRNA. The flags indicate the insertion sites of Tn551. Insertions M2003 and fQIII-8 inactivate femA, whereas insertions fIl-1 and fIII-2 inactivate femB. The position of the 10.5-kb PstI fragment used as a probe in Southern hybridizations is shown as a dotted line. The sizes of the inserts of recombinant plasmids pBBB31 and pBBB64 are shown as heavy lines.

VOL. 36, 1992

nafcillin (10 ,ug/ml) prior to labeling. Separation by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and fluorography of radiolabeled PBPs were carried out as described by Berger-Bachi et al. (5). Isolation of SDS-treated cell walls. SDS-treated cell walls were isolated essentially as described by Potvin et al. (24). Cells from a 1-liter exponential-phase culture of strain BB255 were harvested by centrifugation and washed once in cold distilled water. Cells were broken in five consecutive 2-min intervals with a Beadbeater (Biospec Products, Bartlesville, Okla.). The crude cell walls were harvested by centrifugation and resuspended in 4% (wt/vol) SDS and placed in a water bath with shaking (200 rpm) at 37°C for 90 min and then at 100°C for 30 min. The SDS-treated cell walls were isolated by centrifugation and were washed four times in cold distilled water. Autolysin extraction. One hundred milliliters of LB medium was inoculated with a 2% (vol/vol) inoculum from an overnight culture and allowed to grow with shaking to an optical density at 580 nm of 1.0. The cells were harvested by centrifugation and were resuspended in 1 ml of a solution consisting of 5% sucrose, 2% SDS, and 2% 3-mercaptoethanol; the solution was placed in an Eppendorf tube and frozen overnight at -20°C. The suspension was thawed the next day at 37°C for 1 h and was then centrifuged at high speed in a microcentrifuge. The supernatant containing the autolysins was retained. For growth-phase experiments, cells were grown in 1 liter of LB medium, and the volume of the aliquots was adjusted so that the amount of cells removed at each optical density was similar. The cells were then extracted as described above. The protein concentration was determined with a protein assay (Bio-Rad). Detection of autolysins in cell wall-containing SDS-polyacrylamide gels. All SDS-PAGE assays were carried out in a 12.5% slab gel at 2 mV/cm overnight (9 h). Autolytic banding patterns were determined essentially as described by Jayaswal et al. (15) by adding 1 mg of SDS-treated S. aureus cell walls per ml to the acrylamide solution and brief sonification before polymerization. Transparent bands at the positions of the autolysins were identified after incubation of the gel in 700 ml of 0.05 M K2PO4 (pH 7.2) overnight at 37°C with gentle shaking. Prior to photography, gels were submersed in a solution containing 0.4% methylene blue, 0.01% KOH, and 22% ethanol for 3 min and were washed immediately in cold tap water to enhance the intensities of the autolysin bands. Prestained molecular weight markers (Bio-Rad) were used to estimate the molecular weights of autolytic bands in cell wall-containing gels. Population analysis. Aliquots of overnight cultures were spread onto LB plates containing increasing concentrations of methicillin. CFU were determined after 48 h of incubation at 37°C. RESULTS SmaI and EcoRV mapping of different TnS51 inserts that reduce methicillin resistance. When TnSSI is integrated into the chromosome, it behaves essentially as a stable chromosomal marker. The TnSSI inserts that reduced methicillin resistance in the homogeneously Mcr strain Col were transduced into Mcr recipients BB270 and BB271, selecting for Emr transductants which had lost the ability to grow on methicillin. Mapping of the TnSSl inserts was facilitated in BB270 and BB271, since both strains were Mcr derivatives of NCTC 8325, whose SmaI map of the chromosome has been established (22). Initially, the recipients BB270 and

fem FACTORS IN S. AUREUS

1369

BB271 were thought to be isogenic (2); however, their chromosomal SmaI restriction patterns turned out to be different, whereby BB270 had a smaller SmaI-F fragment than expected (Fig. 2A and C, lanes 1 and 3). This was also true for the SmaI pattern of any fem mutant transductants derived from BB270 (Fig. 2A and C, lanes 1 and 5 to 9) and had to be taken in account when mapping the Tn551 inserts. We identified the position of SmaI-F in BB270 (Fig. 2C); it was truncated by approximately 30 kb compared with BB271 by using phage 411 or an agr probe (which gave results identical to those obtained with 411), both of which are known to map in the SmaI-F fragment (23). The deleted DNA was of an unknown nature and was not phage 411, nor did it involve agr genes. Because of the presence of the mec determinant, the SmaI-G band of all mec-containing strains BB270, BB271, BB308, BB597, BB592, BB593, BB589, and BB591 in Fig. 2A and B (lanes 1, 3, 4, 5, 6, 7, 8, and 9, respectively) moved up by approximately 30 kb compared with the SmaI-G band of the susceptible strain BB255 (Fig. 2A and B, lanes 2). The SmaI-restricted DNAs of the Mcs transductants were probed with TnSSJ (Fig. 2D). One novel insert, f112F, mapped in the SmaI-I fragment (Fig. 2D, lane 9). All other Tn5S51 inserts mapped to the largest SmaI-A fragment (Fig. 2D, lanes 4 to 8), which has a size of more than 673 kb. Four of these inserts (fQ2003, fIII-8, QII-1, and fQIII-2) have been mapped earlier (3), as shown in Fig. 1. The location of the novel SmaI-A insert Q12005 in strain BB589 had not yet been determined. To differentiate its location within the SmaI-A fragment, EcoRV, which does not restrict Tn551, was used to digest chromosomal DNA from the Mcs transductants. First, they were probed with the 10.5-kb PstI fragment covering thefemA andfemB regions (Fig. 3A), and subsequently they were probed with the TnS51 probe (Fig.

3B). Compared with the parent strains BB255, BB270, and BB271 (Fig. 3A, lanes 1, 2, and 3, respectively), the disappearance of the 2.2-kb fragment which was replaced by a 7.4-kb fragment in strains BB308 and BB769 (Fig. 3A, lanes 4 and 5, respectively) indicates the insertion of TnS51 in the femA region. The loss of the 1.2-kb fragment, which was replaced by a 6.4-kb fragment in strain BB592 (Fig. 3A, lane 6), indicates an insertion in thefemB region (3). However, in strains BB589 and BB591 (Fig. 3A, lanes 7 and 8, respectively) containing the Tn551 insertions [12005 and f112F, respectively, no band rearrangement was seen, indicating

that TnS51 had inserted at another location within the chromosome outside of the EcoRV fragments covered by the 10.5-kb PstI probe. Figure 3B shows the same digests as those shown in Fig. 3A but probed with TnSSI; Fig. 3B corroborates the positions of TnSSl. Insertion Q12005 (Fig. 3B, lane 7) appeared in an EcoRV fragment apparently of the same size as the femA insertions; however, this insertion was not within the EcoRV fragments covered by the 10.5-kb PstI probe and, thus, is distinct from those in the femA and femB regions. The insertion f112F in strain BB591 (Fig. 3B, lane 8) migrated in a 13-kb EcoRV fragment within the SmaI-I fragment. We called these two new sites corresponding to insertions [12005 and Q112F femC and femD, respectively. These studies raise the number of factors which affect methicillin resistance in S. aureus to four. PBP assays. In order to determine, first, whether the 30-kb deletion observed in the SmaI-F fragment of the transductants derived from BB270 or, second, whether any of the new fem-inactivating TnS51 insertions might affect the PBP profile or PBP 2' production, PBP assays were performed

1370

BERGER-BACHI ET AL.

ANTimICROB. AGENTS CHEMOTHER.

FIG. 2. Pulsed-field gel electrophoresis of SmaI digests of the chromosomal DNA from different S. aureus mutants. All strains are derivatives of strain NCTC 8325. Lanes 1, Mcr strain BB270 with a 30-kb deletion in SmaI-F; lanes 2, Mcs strain BB255; lanes 3, Mcr strain BB271; lanes 4, strain BB308 carrying fQ2003(femA::Tn551), an insertionally inactivated BB271; lanes 5 to 9 insertionally inactivated transductants in BB270 [lanes 5, strain BB597 with fIII-8(femA::TnS51); lanes 6, strain BB592 with flII-1(femB::Tn551); lanes 7, strain BB593 with f1III-2(femB::Tn551); lanes 8, strain BB589 with f12005(femC::Tn551); lanes 9, strain BB591 with f112F(femD::TnSSI)]; lanes 10, molecular weight markers. (A) Ethidium bromide-stained gel. The positions of the relevant SmaI bands A, F, G, and I in strain BB255 (lane 2) are shown. The SmaI-G band in the different lanes is indicated by a white dot, and the SmaI-F band is indicated by a white arrowhead. (B) Southern blot of the gel shown in panel A probed with mecA, showing that the SmaI-G fragment increased by 30 kb by transduction of the mec determinant. The original positions of SmaI-G (in lane 2) and the position of SmaI-G carrying mec are indicated. (C) Southern blot of the gel shown in panel A probed with 4i11 DNA showing the position of fragment SmaI-F, which was shortened by a spontaneous deletion of approximately 30 kb in strain BB270 and all its transductants. (D) Southern blot of the gel shown in panel A probed with TnS5. The positions of SmaI-A and SmaI-I are indicated.

(Fig. 4). The PBP profile between the two isogenic strains BB271 and BB270, which differed in the 30-kb deletion in SmaI-F, were identical (Fig. 4, lanes 2 and 3). Inactivation of femA either in strain BB308 (lane 4) or in strain BB769 (lane 5) did not alter the production of PBPs compared with that in their corresponding Mcr parent strains BB271 and BB270, respectively. Thus, the PBP profile between these strain pairs is identical and is not influenced by the 30-kb deletion orfemA inactivation. Mutations infemB (lane 6),femC (lane 7), and femD (lane 8) compared with those in the parent strain BB270 (lane 2) also apparently did not alter the production of PBPs, including the low-affinity, mecA-encoded PBP 2'. Autolytic banding patterns. We wanted to determine whether the lowering of whole-cell autolysis observed infem mutants (9, 19) may be due to alteration of autolysin production. In Fig. 5, five major autolysin bands at 140, 110, 87, 59, and 32 kDa were shown to be present in all the various fem mutants when they were compared with their isogenic parent strains. Care was taken to isolate the autolysins from all strains grown to the same optical density, because changes in the autolysin banding patterns of strain BB255 were observed at different phases of growth (data not shown).

During growth, the autolysins at 110, 59, 42, and 32 kDa appeared to increase in activity or quantity as the strains reached the stationary phase of growth. This was demonstrated as greater zones of lysis in the cell wall-containing gels. Population analyses offem mutants. femC and femD inactivation in strain Col converted this homogeneously Mcr strain into a heterogeneously Mcr strain (17). It remained unknown whether inactivation of these factors had any effect on an initially heterogeneously Mcr strain. Population analyses were therefore carried out on each type of fem mutant derived from heterogeneously Mcr strain BB270. Figure 6 shows the population analyses of BB270 and itsfem mutants BB308 (femA), BB592 (femB), BB589 (femC), and BB591 (femD). Tn5Sl insertions in femA and femB appeared to lower the low-level basal resistance as well as the high-level resistance. This effect was greater in femA than in femB mutants. Insertions in the femC and femD regions seemed to lower mainly the low-level basal resistance, while in the higher-level-resistant subpopulations, only the numbers of cells were lowered, yet they were not as greatly diminished as they were in the femA and femB mutants. The overall

VOL. 36, 1992

fem FACTORS IN S. AUREUS

1371

FIG. 4. Autoradiography of PBPs. Lanes: 1, Mcs strain BB255; 2, Mc' strain BB270; 3, Mcr strain BB271; 4, strain BB308 fQ2003 (femA::TnS51); 5, strain BB769 fQ2003(femnA::TnSS1); 6, strain BB592 fiII-1(femB::TnSS); 7, strain BB589 fQ2005(fiemC::TnS51); 8, strain BB591 f112F(femD::Tn551). PBPs were labeled with 3Hpenicillin (lanes a), and PBPs were saturated with 10 ,ug of nafcillin per ml prior to labeling as described in the text (lanes b). The positions of PBPs 1, 2, and 3 and of low-affinity PBP 2' are indicated.

DISCUSSION

FIG. 3. Southern blots of EcoRV digests of the four classes of fem mutants probed with the 10.5-kb PstI fragment covering the femA-femB region (see Fig. 2) (A) and the TnSSl-specific AvaI fragment (B). Lanes: 1, Mcs strain BB255; 2, Mcr strain BB270; 3, Mcr strain BB271; 4, fQ2003(fiemA::TnS51) in strain BB308; 5, fQ2003(femA::TnSS1) in strain BB769; 6, QlII-1(femB::TnSS1) in strain BB592; 7, fl2005(femC::TnSS1) in strain BB589; 8, f112F(fiemD::TnSS1) in strain BB591. The sizes of the bands are indicated (in kilobases).

effect on methicillin resistance was weaker in femC and femD mutants than it was in femrA and femB mutants. Complementation of femA inactivation by pBBB31 and pBBB64 in homogeneously Mcr Col and heterogeneously Mcr BB270. It is not known whether the femA allele from a heterogeneous strain can complement afemA mutation in a homogeneous strain. We therefore tried to complement a femA mutant of strain Col, originally a homogeneously Mcr strain, with plasmid pBBB31 containing the femA region of heterogeneously Mcr BB270. The introduction of pBBB31 into BB403 (a femA mutant in a homogeneous background) or into BB308 (a femA mutant in a heterogeneous background), producing strains BB671 (Fig. 7A) and BB586 (Fig. 7B), respectively, increased resistance in both strains. Similar results were obtained when pBBB64 coding for both femA and femB from BB270 was introduced into the femA mutants, producing strains BB672 (Fig. 7A) and BB664 (Fig. 7B). The complementation of BB308 by pBBB64 seemed slightly better than that with only pBBB31, whereas in BB403, both plasmids had a similar effect. The initial homogeneous resistance of Col was almost but not completely restored.

By using TnS51 insertional mutagenesis, a number of accessory genes named femA factors that map outside the mec determinant were found to be involved with the expression of methicillin resistance (4, 17). Tomasz (29) has also referred to these factors as auxiliary genes (aux). It is probable that the names of these factors will change as their functions become known. We reported here the locations of two additional fem factors in S. aureus, besides the two physically linked genes femA and femB, namely, a third unlinked factor termed femC, which is located on the SmaI-A fragment, and a fourth factor termed femD, which maps on the SmaI-I fragment. The effect that the inactivation of these factors has on the phenotypic expression of methicillin resistance differs in a specific way from factor to factor and can be grouped into two phenotypes. One pheno-

FIG. 5. Autolysin zymograms in cell wall-containing SDS-polyacrylamide gels. The autolysin banding patterns of various fem mutants and their parent strains are compared. Samples were prepared from cells grown to an optical density at 580 nm of 1.0. Lanes: 1, BB255; 2, BB270; 3, BB271; 4, BB308; 5, BB589; 6, BB591; 7, BB592; 8, BB769. Protein was added to each lane at a final concentration of 50 pg. The sizes of the bands are indicated on the left (in kilobases).

1372

ANTimICROB. AGENTS CHEMOTHER.

BERGER-BACHI ET AL.

10

T-

0)

0) 0

0

=J -

C.)

.1

1

10 100 1000 10000 methicillin [jg/ml] FIG. 6. Population analysis of the four different types of fem mutants. Mcr strain BB270 and one representative strain of each insertionally inactivated fem factor, strain BB308 carrying f12003 (femA::TnSSI), strain BB592 carrying fII-1(femB::TnSSl), strain BB589 carrying £12005(femC::TnSS), and strain BB591 with f112F (femD::Tn551), are shown. l, BB270; A, BB308; 0, BB592; *, BB589; *, BB591. type is created by the inactivation of femA or femB and reduces overall resistance, whereby femA has a greater impact than femB. In the second phenotype, represented by femC and femD mutants, only low-level basal resistance is reduced, whereas the amount of highly resistant cells is not as greatly diminished as it is for femA and femB strains. These findings were similar in both homogeneous and heterogeneous Mcr strains (17; this study). This suggests that femA and femB may be more important thanfemC and femD in retaining the number of high-level-resistant cells in heterogeneous Mcr strain BB270 and homogeneous Mcr strain Col. Plasmid pBBB31 and pBBB64 containing the femA and the femA plus femB regions of BB270, respectively, could complement femnA mutants from BB270 and Col; however, they could not complement the femA mutant exactly to the levels of resistance seen in the parent strains. Especially in the femA-inactivated Col strain, complete homogeneous resistance was not obtained. These data indicate that femA and femB genes from a heterogeneous strain can, essentially, complement the femrA mutation in an unrelated homogeneous strain. This does not, however, rule out the fact that differences in the femrA and femB alleles exist between homogeneous and heterogeneous strains. The fact that neither plasmid could completely overcome the chromosomal inactivation of femA in strains BB270 and Col might, on the other hand, be due to a copy number effect of the recombinant plasmid, which might interfere with complementation, or to a polar effect of femA(Q20O3) to important genes downstream of femA and femB. TnSSI inactivation of some of these factors reduces wholecell autolysis and cell wall turnover (9, 19). In femAinactivated strains, this was shown to be due to the lowered glycine content of the peptidoglycan (19). It is conceivable that these events might have been triggered by the inactivation of a PBP or major autolysin. We showed in this study that TnSSI inactivation of all thefem factors affected neither PBP 2' nor major autolysin production found in cell extracts. Therefore, the effects that inactivation of the fem factors

.1

1

.1

1

10 100 methicillin [jg/ml]

1000

10000

1000

10000

10

00

U6-

10

100

methicillin [jig/ml] FIG. 7. Population analysis showing complementation of insertion fQ2003(femA::TnSS1) with either recombinant plasmid pBBB31(femAA) or pBBB64(femAA femB+) in homogeneously or heterogeneously Mcr strains. (A) Population analysis of homogeneously Mcr strain Col and its Q12OO3(femA::TnSS1)-inactivated transductant BB403. Strain BB671 is transductant BB403 complemented with plasmid pBBB31, and strain BB672 is transductant BB403 complemented with pBBB64. El, Col; 0, BB403; *, BB671; 0, BB672. (B) Population analysis of heterogeneously Mcr strain BB270 and its fQ2003(femA::Tn5S1)-inactivated derivative BB308. Strain BB586 is BB308 complemented with pBBB31, and strain BB664 is BB308 complemented with pBBB64. l, BB270; 0, BB308; *, BB586; 0, BB664.

have on cell wall metabolism are presumably primarily an alteration of the substrates of both PBP 2' and the autolysins. Consequently, the altered cell wall turnover and autolysin activity observed might be a secondary effect. It has been reported recently that S. aureus has as many as 21 different autolysins (26). The cell extracts of our strains showed five similar major autolysin bands of 140, 110, 87, 59, and 32 kDa as well as a high number of fainter bands (Fig. 5). S. aureus has been reported to possess three autolytic activities, an N-acetylmuramyl-L-alanine amidase (13, 15, 28), a 3-N-acetylglucosaminidase (6, 32), and a glycine endopeptidase (28). It is also known that cell wall hydrolases

fem FACTORS IN S. AUREUS

VOL. 36, 1992

can be processed to a certain extent and still retain their activities (24) and that the turnover of cell walls in Bacillus subtilis may be regulated by extracellular proteases which act on autolysins (16). Therefore, it is unlikely that S. aureus contains so many enzymes, one represented by each band. We suggest that the fainter autolysin bands may be degradation or modified products of the five major autolysin bands. ACKNOWLEDGMENTS We thank Merck Sharp & Dohme for the generous gift of

[3H]benzylpenicillin. This work was supported by the Swiss National Foundation (grant 32-26279.89). REFERENCES 1. Bachi, B. 1980. Physical mapping of BglI, BglII, PstI, and EcoRI restriction fragments of staphylococcal phage + 11 DNA. Mol. Gen. Genet. 180:391-398. 2. Berger-Bachi, B. 1983. Insertional inactivation of staphylococcal methicillin resistance by TnSSl. J. Bacteriol. 154:479-487. 3. Berger-Bachi, B., L. Barberis-Maino, A. Strassle, and F. H. Kayser. 1989. femA, a host mediated factor essential for methicillin resistance in Staphylococcus aureus: molecular cloning and characterization. Mol. Gen. Genet. 219:263-269. 4. Berger-Bachi, B., and M. L. Kohler. 1983. A novel site on the chromosome of Staphylococcus aureus influencing the level of methicillin resistance: genetic mapping. FEMS Microbiol. Lett. 20:305-309. 5. Berger-Bachi, B., A. Strassle, and F. H. Kayser. 1986. Characterization of an isogenic set of methicillin-resistant and -susceptible mutants of Staphylococcus aureus. Eur. J. Clin. Microbiol. 5:697-701. 6. Biavasco, F., C. Pruzzo, and C. Thomas. 1988. Cloning and expression of the Staphylococcus aureus glucosaminidase in Escherichia coli. FEMS Microbiol. Lett. 49:137-142. 7. Chambers, H. F. 1988. Methicillin-resistant staphylococci. Clin. Microbiol. Rev. 1:173-186. 8. Chambers, H. F., and C. J. Hackbarth. 1987. Effect of NaCl and nafcillin on penicillin-binding protein 2a and heterogeneous expression of methicillin resistance in Staphylococcus aureus. Antimicrob. Agents Chemother. 31:1982-1988. 9. de Jonge, B. L. M., H. de Lencastre, and A. Tomasz. 1991. Suppression of autolysis and cell wall turnover in heterogeneous TnSSI mutants of a methicillin-resistant Staphylococcus aureus strain. J. Bacteriol. 173:1105-1110. 10. Fontana, R. 1985. Penicillin-binding proteins and the intrinsic resistance to 1-lactams in gram-positive cocci. J. Antimicrob.

Chemother. 16:412-416. 11. Goering, R. V., and T. D. Duensing. 1990. Rapid field inversion gel electrophoresis in combination with an rRNA gene probe in the epidemiological evaluation of staphylococci. J. Clin. Micro-

biol. 28:426-429.

12. Hartman, B. J., and A. Tomasz. 1986. Expression of methicillin resistance in heterogeneous strains of Staphylococcus aureus. Antimicrob. Agents Chemother. 29:85-92. 13. Huff, E., C. S. Silverman, N. J. Adams, and W. S. Awkard. 1970. Extracellular cell wall lytic enzyme from Staphylococcus aureus: purification and partial characterization. J. Bacteriol.

103:761-769. 14. Inglis, B., P. R. Matthews, and P. R. Stewart. 1988. The expression in Staphylococcus aureus of cloned DNA encoding

methicillin resistance. J. Gen. Microbiol. 134:1465-1469.

15. Jayaswal, R. K., Y. Lee, and B. J. Wilkinson. 1990. Cloning and expression of a Staphylococcus aureus gene encoding a peptidoglycan hydrolase activity. J. Bacteriol. 172:5783-5788.

1373

16. Jolliffe, L. K., R. J. Doyle, and U. N. Streips. 1980. Extracellular proteases modify cell wall turnover in Bacillus subtilis. J. Bacteriol. 141:1199-1208. 17. Kornblum, J., B. J. Hartman, R. P. Novick, and A. Tomasz. 1986. Conversion of a homogeneously methicillin-resistant strain of Staphylococcus aureus to heterogeneous resistance by TnS51-mediated insertional inactivation. Eur. J. Clin. Microbiol. 5:714-718. 18. Madiraju, M. V. V. S., D. P. Brunner, and B. J. Wilkinson. 1987. Effects of temperature, NaCl, and methicillin on penicillin-binding proteins, growth, peptidoglycan synthesis, and autolysis in methicillin-resistant Staphylococcus aureus. Antimicrob. Agents Chemother. 31:1727-1733. 19. Maidhof, H., B. Reinicke, P. Blumel, B. Berger-Bachi, and H. Labischinsld. 1991. femA, which encodes a factor essential for expression of methicillin resistance, affects glycine content of peptidoglycan in methicillin-resistant and methicillin-susceptible Staphylococcus aureus strains. J. Bacteriol. 173:3507-3513. 20. Maniatis, T., E. F. Fritsch, and J. E. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 21. Matsuhashi, M., M. D. Song, F. Ishino, M. Wachi, M. Doi, M. Inoue, K. Ubukata, N. Yamashita, and M. Konno. 1986. Molecular cloning of the gene of a penicillin-binding protein supposed to cause high resistance to 1-lactam antibiotics in Staphylococcus aureus. J. Bacteriol. 167:975-980. 22. Patel, A. H., T. J. Foster, and P. A. Pattee. 1989. Physical and genetic mapping of the protein A gene in the chromosome of Staphylococcus aureus 8325-4. J. Gen. Microbiol. 135:17991807. 23. Pattee, P. A. 1990. Genetic and physical mapping of the chromosome of Staphylococcus aureus NCTC 8325, p. 163-169. In K. Drlica and M. Riley (ed.), The bacterial chromosome, John Wiley & Sons, Inc., New York. 24. Potvin, C., D. Leclerc, G. Tremblay, A. Asselin, and G. Bellemare. 1988. Cloning, sequencing and expression of Bacillus bacteriolytic enzyme in Escherichia coli. Mol. Gen. Genet. 214:241-248. 25. Reynolds, P. E., and C. Fuller. 1986. Methicillin-resistant strains of Staphylococcus aureus: presence of identical additional penicillin-binding protein in all strains examined. FEMS Microbiol. Lett. 33:251-254. 26. Sugai, M., T. Akiyama, H. Komatsuzawa, Y. Miyake, and H. Suginaka. 1990. Characterization of sodium dodecyl sulfatestable Staphylococcus aureus bacteriolytic enzymes by polyacrylamide gel electrophoresis. 172:6494-6498. 27. Tesch, W., A. Strassle, B. Berger-Bachi, D. O'Hara, P. Reynolds, and F. H. Kayser. 1988. Cloning and expression of methicillin resistance from Staphylococcus epidermidis in Staphylococcus camosus. Antimicrob. Agents Chemother. 32: 1494-1499. 28. Tipper, D. J. 1969. Mechanisms of autolysis of isolated cell walls of Staphylococcus aureus. J. Bacteriol. 97:837-847. 29. Tomasz, A. 1990. Auxiliary genes assisting in the expression of methicillin resistance in Staphylococcus aureus, p. 565-584. In R. P. Novick (ed.), Molecular biology of the staphylococci, VCH Publishers, New York. 30. Tomasz, A., S. Nachman, and H. Leaf. 1991. Stable classes of phenotypic expression in methicillin-resistant clinical isolates of staphylococci. Antimicrob. Agents Chemother. 35:124-129. 31. Ubukata, K., R. Nonoguchi, M. Mitsuhashi, and M. Konno. 1989. Expression and inducibility in Staphylococcus aureus of the mecA gene, which encodes a methicillin-resistant S. aureus specific penicillin-binding protein. J. Bacteriol. 171:2882-2885. 32. Wadstrom, T., and K. Hitsatsune. 1970. Bacteriolytic enzymes from Staphylococcus aureus. Purification of an endo-,B-Nacetylglucosaminidase. Biochem. J. 120:725-744.