Staphylococcus aureus

Transcription

1 ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Dec. 1987, p /87/ $2./ Copyright 1987, American Society for Microbiology Vol. 31, No. 12 Effect of NaCl and Nafcillin on Penicillin-Binding Protein 2a and Heterogeneous Expression of Methicillin Resistance in Staphylococcus aureus HENRY F. CHAMBERS* AND CORINNE J. HACKBARTH Medical Service, San Francisco General Hospital, San Francisco, California 9411 Received 1 April 1987/Accepted 9 September 1987 Expression of methicillin resistance in heterogeneous strains of Staphylococcus aureus is enhanced by 2 to 5% NaCl in the medium and by selection with beta-lactam antibiotics. Resistance is associated with production of a penicillin-binding protein (PBP), PBP 2a, with low affinity for binding beta-lactam antibiotics. Therefore, the effects of NaCl and nafcillin on amounts of PBP 2a produced and its binding affinity were examined and correlated with expression of resistance. Nafcillin-triggered autolysis also was examined. No relationships between the level of resistance expressed and (i) relative amounts of PBP 2a, (ii) inducibility of PBP 2a by nafcillin, or (iii) binding affinity of nafcillin for PBP 2a were found. A protective effect of NaCl for the susceptible subpopulation, corresponding to inhibition of autolysis, was observed for heterogeneous strains. Even in the absence of NaCl, highly resistant cells were relatively tolerant to nafcillin-triggered autolysis. These results support the hypothesis that high levels of resistance require an additional factor besides PBP 2a. This factor may act within the autolytic pathway. Methicillin-resistant staphylococci may be either heterogeneous or homogeneous in their expression of resistance (2, 23). Heterogeneous strains display low levels of resistance and typically are composed of two subpopulations of cells: a majority population of susceptible cells (i.e., cells killed by relatively low concentrations of beta-lactam antibiotic) and a small, highly resistant subpopulation (1 cell in 16) able to grow in the presence of 5,ug or more of methicillin per ml. Various conditions (presence of salt in the media, incubation at 3 C, prior exposure to antibiotics) are known to enhance expression of heterogeneous resistance (23). Supplementation of media with NaCl is recommended when testing staphylococci for methicillin resistance (27). Homogeneous strains display high levels of resistance. These strains are composed of cells uniform in their expression of resistance. Methicillin resistance in Staphylococcus aureus is associated with the production of a novel penicillin-binding protein (PBP), PBP 2a or PBP 2' (5, 11, 13, 15, 3), which has low affinity for binding beta-lactam antibiotics. Because the effects of beta-lactam antibiotics are mediated by binding to PBPs (12), a low-affinity PBP, such as PBP 2a, could confer resistance by continuing to catalyze cell wall synthesis even in the presence of antibiotic. Both homogeneous (5, 13, 15, 3) and heterogeneous (8, 14) strains produce PBP 2a, despite their marked differences in expression of resistance. The mechanisms responsible for this variable expression of resistance are unknown. Hartman and Tomasz (14) have hypothesized that an additional factor besides PBP 2a is involved, on the basis of the observation that expression of resistance in a temperature-sensitive strain changed independently of production of PBP 2a. This additional factor may be acting via the autolytic pathway (6, 14, 21, 3). The following experiments were conducted to explore further the relationships between PBP 2a, autolysis, and expression of methicillin resistance in heterogeneous and * Corresponding author. homogeneous strains. The effects of salt, which, like incubation at 3 C, enhances resistance, and nafcillin, as a representative beta-lactam antibiotic, were examined. MATERIALS AND METHODS Bacterial strains. The strains of S. aureus used for these studies are shown in Table 1. All strains except 1-63, a clinical isolate used to establish experimental endocarditis (9), are,-lactamase negative, which was confirmed by an inability to hydrolyze the chromogenic cephalosporin nitrocephin (BBL Microbiology Systems, Cockeysville, Md.). Two of the strains, 67- and ID-7, have been cured of P-lactamase (3); the others were originally P-lactamase negative. Methicillin-susceptible strains were defined as those from which no colonies grew when an inoculum of 18 CFU was plated onto Trypticase soy agar (TSA) (BBL) containing 5,ug of nafcillin per ml and incubated for 72 h at 37 C. The methicillin-resistant strains were defined by their ability to grow on TSA containing 1,ug of nafcillin per ml. Homogeneously resistant strains were those for which.1% of cells grew on TSA containing 5,ug of nafcillin per ml. Heterogeneously resistant strains were those for which <1% grew on TSA containing 5,ug of nafcillin per ml. Strain 29P was obtained from the American Type Culture Collection, Rockville, Md., and is the same as ATCC Strain Col, strain 27S and its homogeneous methicillinresistant transductant strain 27R, and the heterogeneous strain Du were obtained from Barry Hartman (13, 14). Strain 67- is a,-lactamase-free clone of a heterogeneous,blactamase-positive clinical isolate previously studied in experimental endocarditis models of infection (7, 9), and its PBP 2a is inducible by beta-lactam antibiotics (8). The heterogeneous strain ID-7 is a,b-lactamase-free clone of a reference strain obtained from the College of American Pathologists. Susceptibility studies. MICs were determined by the standard macrotube dilution method (16) in Trypticase soy broth (TSB) (BBL) with and without 4.% NaCl at an inoculum of 1982

2 VOL. 31, 1987 TABLE 1. Susceptibility characteristics of the experimental strains of S. aureus MIC (Lg/ml)b Proportion of resistant cells at Strain Typea following nafcillin concn (p.g/ml)c: - Salt + Salt 1 plg/ml 1,ug/ml 5,ug/ml 27S SUS C.5 C.5 29P SUS C.5 C SUS C.5 C.5 NDd ND ND Du HET C o-3.2 1o HET o-2.4 1o-4.2 1o-5. ID-7 HET R HOM Col HOM a SUS, Methicillin susceptible; HET, heterogeneous methicillin resistant; HOM, homogeneous methicillin resistant. b MICs are for nafcillin and were determined in TSB; - Salt, unsupplemented; + Salt, supplemented with 4% NaCl. c The proportion of resistant cells was determined by plating onto TSA containing one of three concentrations of nafcillin. Values are the number of CFU that grew after 48 h at 37 C at each concentration divided by the number of CFU that grew on drug-free TSA. d ND, Not done. approximately 15 CFU/ml. Tubes were inspected for turbidity after a 24-h incubation at 37 C. The presence of resistant subpopulations was determined by quantitatively subculturing 18 CFU from an 18-h culture in TSB onto TSA containing one of several concentrations of nafcillin. After a 48-h incubation at 37 C, the colonies that had grown at each concentration were counted. Growth curves were determined in TSB, with and without 4% NaCl supplementation. An 18-h culture was diluted 1:1 into 2 ml of fresh broth, and the culture was incubated on a shaking platform at 37 C. The turbidity of the culture was measured by absorbance spectrophotometry (Sequoia- Turner, Mountain View, Calif.) at 57 nm. Nafcillin was added once growth became exponential. Preparation of membrane proteins for PBP assay. Membranes were prepared from 5- to 1,-ml cultures of bacteria grown in TSB, with and without 4% NaCl, in a shaking platform incubator (Labline, Melrose, Ill.) at 16 rpm at 37 C. Exponentially growing cells were harvested at a cell density of 2 x 18 to 5 x 18 CFU/ml, washed once in 5 ml of 1 M NaCl-5 mm Tris hydrochloride-5 mm MgCl2 (ph 7.5), and then suspended in 1 ml of.145 M NaCI-5 mm Tris hydrochloride-5 mm MgCl2 (ph 7.5). To this cell suspension was added (final concentrations) 1,ug each of DNase and RNase (Boehringer Mannheim Biochemicals, Indianapolis, Ind.) per ml, 5,ug of lysostaphin (Sigma Chemical Co., St. Louis, Mo.) per ml, and 4 g of glass beads (diameter, 1,um). The cells were disrupted mechanically in a cell disintegrator (B. Braun, Melsungen, Federal Republic of Germany). The beads were washed, and the supernatant was centrifuged at 8, x g for 1 min at 4 C. Membranes were pelleted from this supernatant by centrifugation at 1, x g for 3 min at 4 C. After two washes the membrane pellet was suspended in 1 mm Tris hydrochloride (ph 7.) to a final protein concentration of 1 mg/ml as assayed by method of Bradford (4). Assay of PBPs. Samples (8,ug) of membrane protein were preincubated either with buffer (direct PBP assay) or nafcillin (competition assay) for 15 min at 37 C. [3H]penicillin (2,ug/ml) was then added for a 15-min incubation at 37 C. Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (17) (1% acrylamide,.13 or.26% bisacrylamide) and stained with Coomassie blue stain. HETEROGENEOUS MRSA 1983 Radiolabeled proteins were detected by fluorography with prefogged Kodak X-Omat film (Eastman Kodak Co., Rochester, N.Y.) (18). Amounts of PBP 2a were measured by scanning densitometry (Quick-Scan, Jr.; Helena Laboratories, Beaumont, Tex.). Autolysin assay. The autolysin assay is a modification of the method of Shockman et al. (24, 25). Assays were performed with exponential-phase cells grown in TSB at 37 C. When the culture had reached an A57 of.25, 1-ml aliquots were distributed into 5-ml Erlenmeyer flasks, which were placed into a 37 C shaking incubator. Nafcillin (ranging from to 1,,ugIml) was added, and the cultures were incubated for 1 h. After 1 h, two portions were removed from each sample and centrifuged at 4, x g for 5 min. One pellet was suspended in buffer (.1 M KCl,.1 M MgC92,.1 M Tris [ph 7.]) (26), and the other was suspended in buffer containing 4% NaCl. Samples were incubated in a 37 C water bath, and A585 readings were obtained hourly. The bacterial culture that was not exposed to nafcillin served as an untreated control. Drug-induced autolysis following the 1-h exposure to nafcillin was calculated as the percent change in absorbance as follows: Ai of treated cells/a of treated cells x 1% Ai of control cells/a of control cells where Ai is the absorbance at hour i following the 1 h of drug exposure and Ao is the absorbance at time immediately after cells were resuspended in buffer. RESULTS Effect of NaCi on expression of resistance. As expected, MICs of nafcillin for heterogeneous strains were higher in TSB plus 4% NaCl than in unsupplemented TSB (Table 1). MICs for susceptible and homogeneously resistant strains were similar for TSB and TSB plus 4% NaCl. For heterogeneous strains the proportion of CFU that could grow on TSA containing nafcillin was increased by the addition of 4% NaCl (Table 2). For example, for strain 67-, a 1-fold reduction in CFU was observed at 1,ug of nafcillin per ml with TSA. However, with TSA plus 4% NaCl, this reduction in CFU did not occur until 5,ug of nafcillin per ml was present. Plating experiments were performed to examine whether TABLE 2. Number of CFU expressing resistance at each of several concentrations of nafcillin in TSA with and without 4% NaCl supplementation Strain 4o% NaCl Log1o CFUa at following nafcillin concn (,ug/ml): added P Du ID R Col a Underlined values indicate the concentration at which the pattern of resistance within the colony shifted from heterogeneous to homogeneous.

3 1984 CHAMBERS AND HACKBARTH the increased resistance with NaCl was due to emergence of the resistant subpopulation or to protection of the susceptible cells. For each heterogeneous strain, five colonies were picked randomly from the TSA containing the lowest nafcillin concentration (the underlined values in Table 2) that caused a 1-fold or greater reduction in CFU. For example, for strain 67-, five colonies were picked from the plate containing 1,ug of nafcillin per ml. These were individually inoculated into TSB and incubated for 18 h at 37 C. Samples from each culture were quantitatively subcultured onto TSA containing 5,ug of nafcillin per ml and incubated at 37 C for 48 h. For each heterogeneous strain, one or more of the five colonies subcultured expressed high-level homogeneous resistance. In contrast, all colonies picked from TSA at the next lower nafcillin concentrations (i.e., those that caused less than a 1-fold reduction in CFU) expressed heterogeneous resistance like that of the parent strain. The experiment was repeated with colonies picked from TSA plus 4% NaCl. Only at 5,ug/ml did colonies begin to express homogeneous resistance. Thus, NaCl enhanced the expression of resistance by protection of the more susceptible cells, not by emergence of highly resistant cells. On the contrary, at the lower nafcillin concentrations, NaCl prevented selection of the highly resistant subpopulation. This protective effect of NaCl for susceptible cells is similar to that described for incubation at 3 C for temperature-sensitive heterogeneous strains (14). For the stable homogeneous strains 27R and Col, 4% NaCl in the TSA did not enhance resistance. In fact, at 5 pug of nafcillin per ml, strain Col grew less well in the presence of 4% NaCl. Expression of resistance in homogeneous strains was not increased by passage on TSA containing nafcillin. The effect of NaCl on nafcillin-induced lysis in growing cells was examined. In TSB, cells of heterogeneous strains began to lyse within 2 h of exposure to nafcillin (Fig. 1A). Regrowth at 24 h corresponded to emergence of the highly resistant subpopulation. In the presence of 4% NaCl, lysis was not observed after addition of nafcillin (Fig. 1B and C). However, once the culture was backdiluted into TSB with- ANTIMICROB. AGENTS CHEMOTHER. out NaCl but containing even 1 p,g of nafcillin per ml, lysis occurred. These experiments also demonstrate that NaCl enhanced resistance by protecting susceptible cells. Removal of NaCl from the medium resulted in the loss of the protective effect. Similar results have been reported for temperature effects in the temperature-sensitive heterogeneous strain Du (14). Effects of NaCl on PBP 2a. PBP 2a is inducible in some strains (8, 22, 28). The possibility that NaCl enhanced resistance by induction of PBP 2a was investigated. PBP 2a is inducible by nafcillin in strain 67-, yet 4% NaCl did not induce PBP 2a in this strain. Membranes prepared from cells grown either with or without NaCl supplementation contained similar amounts of radiolabeled PBP 2a (Fig. 2; Table 3). Likewise, for the heterogeneous strains Du and ID-7, neither radiolabeled PBP 2a (Table 3) nor Coomassie blue-stained protein at the position of PBP 2a (Fig. 3) was increased in the presence of NaCl. Similar results have been reported for strain Du, which showed no alterations in amounts of PBP 2a with changes in temperature (14). Because NaCl was not an inducer of PBP 2a, the possibility that NaCl enhanced resistance by alteration of nafcillin binding to PBP 2a was studied. NaCI had no effect on the binding of nafcillin to PBP 2a (Fig. 4). Relationship between amounts of PBP 2a and expression of resistance. A correlation between increased amounts of PBP 2a and increased expression of resistance was examined in two ways. (i) Expression of resistance was correlated with amounts of PBP 2a produced in the inducible strain 67-. (ii) The amount of PBP 2a produced by highly resistant cells selected from a heterogeneous strain by passage in antibiotic was compared with that produced by the relatively susceptible parent. There was no correlation between amounts of PBP 2a produced and the level of resistance expressed. Figure 5 shows a gel of four membrane preparations of strain 67- grown in TSB plus 4% NaCl. In this strain, the amounts of PBP 2a produced differed depending upon whether the P-lactamase-encoding plasmid was present (lanes 1 and 2) or had been eliminated (lanes 3 and 4) and whether nafcillin was E c CN ' c UL- -o E a!9 - - E NO ' - -D Hours Hours Hours FIG. 1. Growth curves for the heterogeneous strain 67-. Cell density was measured by absorbance spectrophotometry at 57 nm. Treated cultures were dosed with 1,ug of nafcillin per ml as indicated by the asterisk. The arrows indicate backdilution steps. Symbols:, TSB; *, TSB plus 4% NaCl; A, TSB plus 1,ug of nafcillin per ml; A, TSB plus 4% NaCl plus 1,ug of nafcillin per ml; O, 1,ug of nafcillin per ml in TSB;, 5,ug of nafcillin per ml in TSB. (A) Effect of nafcillin upon growth in TSB. (B) Effect of pretreatment of strain 67- (cured of P-lactamase) for 9 min with nafcillin in TSB plus 4% NaCl upon subsequent growth after backdilution into fresh TSB both with and without nafcillin. (C) Same as panel B, except that results are for the,-lactamase-producing original isolate of 67-.

4 VOL. 31, 1987 HETEROGENEOUS MRSA 1985 FIG. 2. Fluorograph of [3H]penicillin-radiolabeled PBPs from membrane preparations of strain 67- grown in TSB (lanes 1), TSB plus 4% NaCl (lanes 2), TSB plus 1,ug of nafcillin per ml for 45 min (lanes 3), and TSB plus 4% NaCl plus 1,ug of nafcillin per ml for 45 min (lanes 4). The numbers at the left indicate PBPs. (A) Direct assay with 2,ug of [3H]penicillin per ml. (B) Competition assay with a 15-min preincubation with 1,ug of nafcillin per ml followed by addition of [3H]penicillin. TABLE 3. Effect of salt and nafcillin on the proportion of cells expressing high-level resistance to nafcillin and upon the amount of radiolabeled PBP 2a in fluorographs, as measured by scanning densitometry Growth conditionsa Strain Nafcillin Proportion of Relative density 4% Salt (,Lg/ml, ~~~~~highly CFU at resistant of PBP 5,ug/ml 2ab time) , 6 min l , 36 h NNc , 6 min Du , 36 h ID , 18 h ID-7-NNC - 1, 18 h R , 18 h , 18 h.7.9 a Conditions under which the cultures for preparation of membranes were grown, indicating whether 4% NaCl was added, the concentration of nafcillin present, and the duration of exposure at this concentration. b The ratio of density of radiolabeled PBP 2a in the experimental sample to the density in a sample prepared from the parent strain grown in TSB without 4% NaCl or nafcillin. PBP 2a in all samples was labeled in a competition assay with 1 ±g of nafcillin per ml to saturate high-affinity PBPs, particularly PBP 2, which comigrates with 2a because of its similar molecular weight. c -NN indicates a highly resistant subpopulation selected by two passages in 5 i.g of nafcillin per ml. FIG. 3. Coomassie blue-stained sodium dodecyl sulfate-polyacrylamide gel of proteins in membrane preparations of (A) strain Du and (B) strain ID-7. Molecular mass standards are 92.5, 66.2, 45, and 31 kilodaltons. Cultures were grown in TSB (lanes 1), TSB plus 4% NaCl (lanes 2), and TSB plus 4% NaCl plus 1,tg of nafcillin per ml for 9 min (lanes 3). The position of PBP 2a is indicated by the black dash on the left. For strain Du the ratio of density, measured by scanning densitometry, at the position of PBP 2a in lane 2 compared with lane 1 is 1.2; for lane 3 compared with lane 1, it is 1.1. For strain ID-7, these values are 1. and 1.. present (lanes 2 and 4) or not (lanes 1 and 3). The largest amounts of PBP 2a were induced in the P-lactamase-negative strain grown in nafcillin (lane 4). Examination of Fig. 1B and C (the cultures for which were grown under the same 1.2r co d.8 a-+ C o.4 - * = , Concentration of Nafcillin (Lg/mI) FIG. 4. Effect of 4% NaCl on binding of nafcillin to PBP 2a. Membrane preparations of strain 67- grown in TSB plus 4% NaCl plus 1,ug of nafcillin per ml for 9 min were used for these assays. Ten portions of 15.g of membrane protein were dispensed into borosilicate tubes. Half the samples were suspended in 1 mm Tris hydrochloride (ph 7.) buffer plus 4% NaCl (), and half were suspended in an equal volume of NaCl-free buffer (). PBPs were assayed in competition with, 1, 1, 1, and 1,,ug of nafcillin per ml. The relative density is the ratio of the fluorographic density of radiolabeled PBP 2a after nafcillin competition assay to the density of PBP 2a in the direct assay (i.e., no nafcillin in the preincubation step).

5 1986 CHAMBERS AND HACKBARTH ANTIMICROB. AGENTS CHEMOTHER. 11$ S2 CNJ cl o CL. va >) 1= 31 FIG. 5. Coomassie blue-stained sodium dodecyl sulfate-polyacrylamide gel of proteins of membrane preparations of strain 67- grown in 4% NaCl. Values at left are molecular masses in kilodaltons. Membranes were prepared from both the,-lactamase-positive strain (67-) and this strain cured of its P-lactamase. Lanes 1 and 2 are samples of the original P-lactamase-producing strain grown in no nafcillin (lane 1) or 1,ug of nafcillin per ml (lane 2) for 9 min. Lanes 3 and 4 are the strain cured of,b-lactamase and grown in no nafcillin (lane 3) and 1 p.g of nafcillin per ml (lane 4) for 9 min. PBP 2a corresponds to the dark band at approximately 78 kilodaltons that is above the 66-kilodalton marker. The band above 31 kilodaltons in lane 2 probably is,-lactamase. experimental conditions used in the preparation of the membrane samples depicted in Fig. 5) shows that despite relatively large amounts of PBP 2a, the culture still lysed once it was backdiluted into TSB containing as little as 1 p,g of nafcillin per ml. A correlation between amounts of PBP 2a produced by a strain and the proportion of highly resistant CFU was sought, but none was found (Table 3). For example, strain Du grown for 36 h in TSB containing 5 p.g of nafcillin per ml contained 46% highly resistant cells, yet amounts of radiolabeled PBP 2a were not significantly different from that for a culture in which.3% of cells expressed high-level resistance. Similar results were obtained for strain ID-7-NN, derived from ID-7 by passage on 5,ug of nafcillin per ml. After growth for 18 h in 1,ug of nafcillin per ml, 62% of cells were highly resistant, but the amounts of PBP 2a were no different from that in the parent strain, in which.5% of cells were highly resistant. These results contrast with those reported previously for the strain pair 67- and 67-NN (8), in which the latter, a homogeneous derived strain, produced more PBP 2a than the parent did. Relationship between nafcillin binding to PBP 2a and expression of resistance. Differences in binding of nafcillin to PBP 2a between homogeneous and heterogeneous strains were not observed (Fig. 6). For example, the heterogeneous strains Du and 67- and the homogeneous strain 27R produced PBP 2a that had identical profiles for binding nafcillin. Likewise, heterogeneous-homogeneous strain pairs had identical profiles, even though the homogeneous derivative was composed of a 1,-fold-higher number of highly resistant cells. Effect of NaCl and nafcillin on whole-cell lysis. In the Concentration of Nafcillin (jig/ml) FIG. 6. Competition assay of PBP 2a versus nafcillin in heterogeneous and homogeneous strains. The relative density is calculated as described in the legend to Fig. 4. Symbols: O, strain Du grown in TSB plus 4% NaCl; *, strain 27R grown in TSB plus 4% NaCl; *, strain 67-NN grown in TSB plus 1,ug of nafcillin per ml for 9 min;, strain 67- grown in TSB plus 4% NaCl; A, strain ID-7 grown in TSB plus 4% NaCl; A, strain ID-7-NN grown in TSB. absence of NaCl, heterogeneous strains lysed after exposure to 1,ug of nafcillin per ml (Table 4). Interestingly, the heterogeneous strains Du and 67- were the most lytic. Strain ID-7, which is more resistant than strains 67- or Du but less resistant than strain 27R (Tables 2 and 3), was less lytic than the first two and more lytic than the last. Homogeneous cells derived from heterogeneous strains differed from the parents in that lysis occurred only at higher nafcillin concentrations (e.g., 1,,ug/ml). The heterogeneous-homogeneous strain pair 67- and 67-NN is representative of the pattern of lysis observed for heterogeneous strains (Fig. 7). Lysis was triggered at 1,ug of nafcillin per ml in the parent. The derived, highly resistant strain did not lyse until a concentration of 1,,ug/ml. Dose-ranging studies of strain 67-NN, performed to define more precisely the nafcillin concentration that triggered lysis, showed that maximal lysis was triggered at concentrations of 25 to 5 p,g/ml. At 1,ug/ml and below, the rate TABLE 4. Autolytic activity as determined by densitometry measurements at A585 % Decrease in A585 after 6 h in autolysin buffere at following nafcillin concn (,ug/ml): Strain 1 FLg/ml 5 p.g/ml 1, p.g/ml - Salt + Salt - Salt + Salt - Salt + Salt 27S P NNb Du Du-NNb ID ID-7-NNb R a Autolysin buffer was unsupplemented (- Salt) or supplemented with 4% NaCl (+ Salt). bthe derived, highly resistant strains were grown overnight in TSB plus 1,ug of nafcillin per ml. On the day of the experiment, nafcillin was removed by centrifugation and cells were suspended in 1 ml of fresh TSB.

6 VOL. 31, 1987 LO ' 6 \ Hours FIG. 7. Autolytic activity in whole-cell lysis assays. Symbols: strain 67-; , strain 67-NN; A, treatment with 1,ug of nafcillin per ml and assay in autolysin buffer; A, treatment with 1,ug of nafcillin per ml and assay in autolysin buffer plus 4% NaCl; *, treatment with 1,,ug of nafcillin per ml and assay in autolysin buffer. of lysis was similar to that of unexposed cells. At 25 plg/ml and above, cell lysis was like that observed at 1, plg/ml. NaCl inhibited lysis in all strains over the range of nafcillin concentrations (Table 4). Lysis in susceptible strains also was inhibited by NaCl. DISCUSSION Methicillin resistance, at least in part, is mediated by PBP 2a. Although both heterogeneous and homogeneous strains produce PBP 2a, expression of resistance differs markedly between these two types. Factors affecting the expression of resistance were examined for possible effects on production of PBP 2a, and vice versa. The fact that PBP 2a can be inducible by beta-lactam antibiotics (8, 22, 28) afforded an opportunity to test the relationship between production of PBP 2a and expression of resistance. Inducibility of PBP 2a depends partly on the presence of a plasmid encoding P-lactamase. Ubukata et al. (28) reported that PBP 2a production was constitutive in,-lactamasenegative strains and inducible in,b-lactamase-positive strains. Consistent with this finding, in cells of strain 67- not cured of 13-lactamase, PBP 2a was detectable in Coomassie blue-stained membranes only if they had been exposed to nafcillin (compare Fig. 5, lanes 1 and 2). However, strain 67- retained inducibility of PBP 2a even if cured of - lactamase: the largest amounts of PBP 2a were produced by P-lactamase-negative cells grown in nafcillin (Fig. 5, lane 4). The inducibility of PBP 2a in strain 67- resembled that of the methicillin-resistant strains examined by Rossi et al. (22). The explanation of the differences in inducibility of PBP 2a among strains was not investigated. Perhaps regulatory genes controlling penicillinase inducibility are still present, either on another plasmid or in chromosomal DNA, even though r-lactamase is not produced. Whatever the explanation, the fortuitous variability in amounts of PBP 2a in strain 67- as a function of the presence of the 1-lactamase plasmid and exposure to nafcillin permitted us to test the relationship between inducibility HETEROGENEOUS MRSA 1987 of PBP 2a and amounts produced and heterogeneous expression of resistance. Strain 67- lysed even when large amounts of PBP 2a were induced. In addition, strains Du and ID-7 were heterogeneous, but not inducible. Addition of 4% NaCl enhanced expression of resistance but did not alter amounts of PBP 2a produced. Thus, inducibility of PBP 2a does not mediate differences in expression of resistance. In the previous report describing the inducibility of PBP 2a (8), a homogeneously resistant strain, 67-NN, derived by nafcillin selection from strain 67-, produced larger amounts of PBP 2a than did the parent in TSB unsupplemented with NaCl but containing nafcillin. We speculated this might enable strain 67-NN to grow in relatively high concentrations of nafcillin. In experiments with resistant homogeneous subpopulations derived from strains Du and ID-7, high levels of resistance could be expressed without alterations in amounts of PBP 2a. Likewise, strain 67- grown in TSB plus 1 plg of nafcillin per ml and 4% NaCl (to prevent lysis and allow cells to grow) produced large amounts of PBP 2a, yet lysed when backdiluted into TSB containing 1 plg of nafcillin per ml. Thus, it must be concluded that differences in amounts of PBP 2a produced cannot account for heterogeneous expression of resistance. The increase in PBP 2a in strain 67-NN exposed to nafcillin probably is a reflection of its inducibility during cell growth and is not the explanation for growth. Because strain 67- lyses in the absence of NaCl, a relatively smaller amount of PBP 2a is induced, and not vice versa. Alterations in binding affinity of PBP 2a for nafcillin were sought as an explanation for differences in expression of resistance. Although 4% NaCl enhanced the expression of resistance, it did not affect the binding affinity of nafcillin for PBP 2a. Likewise, homogeneous and heterogeneous patterns of resistance were not accompanied by alterations in the binding affinity of nafcillin for PBP 2a. Studies of nafcillin-triggered autolysis implicated the participation of this pathway in the expression of resistance. NaCl inhibited autolysis in whole cells, just as has been reported for isolated cell walls (29), and heterogeneous strains differed in their pattern of lysis from homogeneous strains, whether naturally occurring or derived by antibiotic selection. The two most heterogeneous strains, 67- and Du, were also the most lytic, even compared with fully susceptible strains. Whether this is a general property of heterogeneous strains cannot be determined from these studies. These studies provide confirmatory evidence to support the hypothesis first proposed by Hartman and Tomasz (14): full expression of methicillin resistance requires the presence of an additional factor besides PBP 2a. As early as 1972, determinants physically unlinked to the methicillin resistance determinant were found to influence the expression of resistance (1). Berger-Bachi and Kohler (1, 2) reported that the erythromycin-resistant transposon Tn55I inserted in chromosomal DNA at a site distant from the methicillin resistance determinant suppressed resistance. Insertion of Tn551 into cells of the homogeneous strain Col was found to produce heterogeneous resistance without apparent alterations in PBP 2a production (A. Tomasz, J. Kornblum, and B. J. Hartman, Program Abstr. 26th Intersci. Conf. Antimicrob. Agents Chemother., abstr. no. 519, 1986). The nature of this additional factor remains to be characterized, but it may act via the autolytic pathway, as others have speculated (6, 14, 21, 3) and as our studies suggest. In heterogeneous strains, this additional factor may be absent or its production may be repressed (14). Strains may

7 1988 CHAMBERS AND HACKBARTH appear susceptible because autolysis is readily triggered at low drug concentrations. Since lysis is triggered at concentrations that bind little or no PBP 2a, the event probably is triggered as a consequence of drug binding to PBPs other than PBP 2a. Like incubation at 3 C (14), the effect of NaCl is that of protecting the susceptible population of cells. For NaCl, the effect appears to result from its inhibition of antibiotictriggered autolysis. This effect, although not specific for resistant cells, presumably enhances resistance if PBP 2a is present, because the cell wall continues to be synthesized. In the absence of NaCl, the susceptible subpopulation lyses and dies, but a subpopulation of homogeneously resistant cells is selected out. Perhaps these cells survive because they have acquired the necessary additional factor by antibiotic selection of a mutation(s) in the autolytic pathway. Studies of the highly resistant derived strains, 67-NN, Du- NN, and ID-7-NN, showed that autolysis was triggered at concentrations of nafcillin that began to saturate PBP 2a (i.e., 25 to 1,,ug/ml). Perhaps autolysis in these antibiotic-selected strains has become dissociated from highaffinity PBPs and now is triggered by drug binding to low-affinity PBP 2a. The stable homogeneous strains, Col and 27R, were poorly lytic whether exposed to antibiotic or not (as has been reported for another homogeneous strain [21]). These strains may be tolerant, as well as resistant, like some strains of penicillin-resistant pneumococci (19). Naturally occurring homogeneously resistant strains may express homogeneous resistance because they produce PBP 2a and constitutively produce the required additional factor. The identification of this additional factor or factors probably will increase our understanding of the mechanism of action of beta-lactam antibiotics. LITERATURE CITED 1. Berger-Bachi, B Insertional inactivation of staphylococcal methicillin resistance by Tn551. J. Bacteriol. 154: Berger-Bachi, B., and M. L. Kohler A novel site on the chromosome of Staphylococcus aureus influencing the level of methicillin resistance: genetic mapping. FEMS Microbiol. Lett. 2: Bouanchaud, D. H., M. R. Scavizzi, and Y. A. Chabbert Elimination by ethidium bromide of antibiotic resistance in enterobacteria and staphylococci. J. Gen. Microbiol. 54: Bradford, M. M A rapid and sensitive method for quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: Brown, D. F. J., and P. E. Reynolds Intrinsic resistance to beta-lactam antibiotics in Staphylococcus aureus. FEBS Lett. 122: Bruns, W., H. Keppler, and R. Baucks Suppression of intrinsic resistance to penicillins in Staphylococcus aureus by polidocanol, a dodecyl polyethyleneoxoid ether. Antimicrob. Agents Chemother. 27: Chambers, H. F., C. J. Hackbarth, T. A. Drake, M. G. Rusnak, and M. A. Sande Endocarditis due to a methicillinresistant Staphylococcus aureus in rabbits: expression of resistance to beta-lactam antibiotics in vivo and in vitro. J. Infect. Dis. 149: Chambers, H. F., B. J. Hartman, and A. Tomasz Increased amounts of a novel penicillin-binding protein in a strain of methicillin-resistant Staphylococcus aureus exposed to nafcillin. J. Clin. Invest. 76: Chainbers, H. F., and M. A. Sande Teicoplanin versus nafcillin and vancomycin in the treatment of experimental endocarditis caused by methicillin-susceptible or -resistant Staphylococcus aureus. Antimicrob. Agents Chemother. 26: Cohen, S., C. J. Gibson, and H. M. Sweeney Phenotypic ANTIMICROB. AGENTS CHEMOTHER. suppression of methicillin resistance in Staphylococcus aureus by mutant noninducible penicillinase plasmids. J. Bacteriol. 112: Georgopapadakou, N. H., and F. Y. Liu Penicillin-binding proteins in a Staphylococcus aureus strain resistant to specific beta-lactam antibiotics. Antimicrob. Agents Chemother. 18: Georgopapadakou, N. H., S. A. Smith, and D. P. Bonner Binding of beta-lactam antibiotics to penicillin-binding proteins of Staphylococcus aureus and Streptococcus faecalis: relation to antibacterial activity. Antimicrob. Agents Chemother. 22: Hartman, B. J., and A. Tomasz Low-affinity penicillinbinding protein associated with beta-lactam resistance in Staphylococcus aureus. J. Bacteriol. 158: Hartman, B. J., and A. Tomasz Expression of methicillin resistance in heterogeneous strains of Staphylococcus aureus. Antimicrob. Agents Chemother. 29: Hayes, M. V., N. A. C. Curtis, A. W. Wyke, and J. B. Ward Decreased affinity of a penicillin-binding protein for betalactam antibiotics in a clinical isolate of Staphylococcus aureus resistant to methicillin. FEMS Microbiol. Lett. 1: Jones, R. N., A. L. Barry, T. L. Gavan, and J. A. Washington Susceptibility tests: microdilution and macrodilution broth procedures, p In E. H. Lennette, A. Balows, W. J. Hausler, Jr., and H. J. Shadomy (ed.), Manual of clinical microbiology, 4th ed. American Society for Microbiology, Washington, D.C. 17. Laemmli, U. K Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227: Laskey, R. A., and A. D. Mills Quantitative film detection of 3H and "4C in polyacrylamide gels by fluorography. Eur. J. Biochem. 56: Liu, H., and A. Tomasz Penicillin tolerance in multiply drug-resistant natural isolates of Streptococcus pneumoniae. J. Infect. Dis. 152: Matthews, P. R., and P. R. Stewart Resistance heterogeneity in methicillin-resistant Staphylococcus aureus. FEMS Microbiol. Lett. 22: Qoronfleh, M. W., and B. J. Wilkinson Effects of growth of methicillin-resistant and -susceptible Staphylococcus aureus in the presence of beta-lactams on peptidoglycan structure and susceptibility to lytic enzymes. Antimicrob. Agents Chemother. 29: Rossi, L., E. Tonin, Y. R. Cheng, and R. Fontana Regulation of penicillin-binding activity: description of a methicillin-inducible penicillin-binding protein in Staphylococcus aureus. Antimicrob. Agents Chemother. 27: Sabath, L. D., and S. J. Wallace Factors influencing methicillin resistance in staphylococci. Ann. N.Y. Acad. Sci. 182: Shockman, G. D., M. J. Conover, J. J. Kolb, P. M. Phillips, L. S. Riley, and G. Toennies Lysis of Streptococcus faecalis. J. Bacteriol. 81: Storch, G. A., and D. J. Krogstad Antibiotic induced lysis of enterococci. J. Clin. Invest. 68: Takebe, I., H. Singer, E. Wise, and J. T. Park Staphylococcus aureus H autolytic activity: general properties. J. Bacteriol. 12: Thormsberry, C., and L. K. McDougal. i983. Successful use of broth microdilution in susceptibility tests for tnethicillinresistant (heteroresistant) staphylococci. J. Clin. Microbiol. 18: Ubukata, K., N. Yamashita, and M. Koono Occurrence of a beta-lactam-inducible penicillin-binding protein in methicillinresistant staphylococci. Antimicrob. Agents Chemother. 27: Wilkinson, B. J., K. J. Dorian, and L. D. Sabath Cell wall composition and associated properties of methicillin-resistant Staphylococcus aureus strains. J. Bacteriol. 136: Wyke, A. W., J. B. Ward, and M. V. Hayes Synthesis of peptidoglycan in vivo in methicillin-resistant Staphyloccus aureus. Eur. J. Biochem. 127: