513

Vol. 158, No. 2

JOURNAL OF BACTERIOLOGY, May 1984. p. 513-516 0021-9193/84/050513-04$02.00/0 Copyright t 1984, American Society for Microbiology

Low-Affinity Penicillin-Binding Protein Associated with r-Lactam Resistance in Staphylococcus aureus BARRY J. HARTMAN* AND ALEXANDER TOMASZ Division of Infectious Diseases, Cornell University Medical Center, and the Rockefeller University, New York, New York Received 31 October 1983/Accepted 23 February 1984

Methicillin resistance in Staphylococcus aureus has been associated with alterations in the penicillinbinding proteins (PBPs). An intriguing property of all methicillin-resistant staphylococci is the dependence of resistance on the pH value of the growth medium. Growth of such bacteria at pH 5.2 completely suppressed the expression of methicillin resistance. We have examined the PBP patterns of methicillinresistant staphylococci grown at pH 7.0. We detected a high-molecular-weight PBP (PBP-2a; approximate size, 78,000 daltons) that was only present in the resistant bacteria but not in the isogenic sensitive strain. In cultures grown at pH 5.2, the extra PBP was not detectable.

*

MATERIALS AND METHODS Bacteria. An isogenic pair of strains 27 and 27R was kindly supplied by Richard P. Novick, New York City Public Health Research Institute. Methicillin resistance of a clinical isolate of strain 592 was introduced into methicillin-sensitive recipient strain 27 by transduction with phage 80ot. The methicillin-resistant strain Col was supplied by Leon Sabath, University of Minnesota and originated from K. G. Dyke, Strain 209-P (ATCC 4538-P) was a second sensitive control strain. Relevant characteristics of these strains are described in Table 1. Culture growth was monitored by measuring optical density or by light scattering by a Coleman nephocolorimeter (Coleman Instruments, Oak Brook, Ill.) (9). MICs. MICs were determined by the tube dilution method in tryptic soy broth (Difco Laboratories, Detroit, Mich.), starting with 105 CFU/ml. Turbidity was read after 48 h of incubation at 37°C. MICs determined by the agar dilution method at 30°C on tryptic soy agar were within two- to fourfold of the values found by the tube dilution method. Labeling of PBPs in growing bacteria. In our experiments we used whole living cells. Bacterial strains were stored at -70°C. Cultures were grown to late logarithmic phase in tryptic soy broth at pH 7.0 or 5.2 at 37°C with vigorous shaking. These were then rediluted by placing 1.0 ml of bacteria into 100 ml of fresh warm tryptic soy broth at the appropriate pH value and grown again to the midlogarithmic phase of growth. The bacteria were centrifuged at 10,000 rpm at 0°C for 15 min, and the pellets were resuspended in 1.0 ml of clean tryptic soy broth (1OOX concentration) at the

appropriate pH. Samples of cells (25 RI) were incubated with various concentrations of [3H]benzylpenicillin (ethylpiperidinium salt) with a specific activity of 25 Ci/mmol (14). Immediately before the [3H]benzylpenicillin was used, the acetone solvent was evaporated and replaced by an equal volume of distilled water. The samples were incubated with [3H]benzylpenicillin at 37°C for 10 min and then were boiled for 2 min. This heat inactivation step considerably improved the resolution of the gels. Next, 25 [LI of 0.1 M potassium phosphate buffer (pH 7.6) was added to all tubes to allow lysis by lysostaphin (100 pLg/ml), and the samples were

Corresponding author.

513

Downloaded from jb.asm.org by on March 19, 2009

affinity PBP-2a was no longer detectable in bacteria grown at the low pH value. Identical observations were noted when a resistant clinical isolate was used.

Intrinsic beta-lactam resistance associated with penicillinbinding protein (PBP) alterations appears to be a newly recognized mechanism present in both laboratory and clinical isolates of resistant strains in a variety of different bacteria (3, 5, 7, 12, 19, 21). The mechanism of resistance appears to be due to an alteration in the target proteins (PBPs) either in their electrophoretic patterns (8) or in their affinities for the beta-lactam antibiotics (2, 8-10). In at least some cases, it was possible to demonstrate by genetic techniques that the observed PBP alterations were causally related to the decreased susceptibility of the bacteria to the antibiotic (3, 5, 21). This is particularly important for methicillin-resistant Staphylococcus aureus strains, since they have become important nosocomial and community-acquired infectious agents (4, 17, 20). A peculiar property of methicillin-resistant staphylococci is the rapid loss of phenotypic resistance by a shift of the pH value of the culture medium from the physiological pH value of 7.0 to pH 5.2 (15, 16). This pH-dependent drop in the MIC appears to be shared by all methicillin-resistant strains (see Table 1). The mechanism of this phenomenon is unknown. In a previous study (9), we demonstrated that resistant strains grown at both pH 7.0 and 5.2 had the same binding pattern of the high-molecular-weight PBPs 1, 2, and 3. In this communication, we examined the PBPs of a pair of isogenic methicillin-resistant and methicillin-sensitive S. aureus strains as well as a resistant clinical isolate. With improvements in our PBP binding and electrophoretic technique, we could demonstrate the presence of an additional PBP with a low affinity for beta-lactam antibiotics in cultures grown at pH 7.0. No such extra band was observable in the isogenic sensitive strain. Labeling of the PBPs in live growing bacteria indicated that saturation of the extra band (PBP2a) required penicillin and methicillin concentrations in the vicinity of the MIC for these antibiotics, whereas all other PBPs could be saturated below the MIC. We also applied our PBP binding and gel system to the reexamination of the PBPs of the same isogenic pair of methicillin-resistant and methicillin-sensitive S. aureus strains grown at either pH 7.0 (allowing expression of resistance) or pH 5.2 (at which the bacteria are phenotypically sensitive to beta-lactams). It was found that the low

514

HARTMAN AND TOMASZ

J. BACTERIOL.

TABLE 1. S. aureus strains MIC (,ug/ml) of:

Methicillin pH 7.0 pH 5.2 0.8 0.2 1.6 0.4 625 3.1 1250 1.6

Straina

Benzylpenicillin pH 7.0 0.2 0.025 10 31

pH 5.2 0.2 0.025 0.025 0.025

209-P 27 27R Col a All strains were beta-lactamase negative by the Nitrocefin assay

(13).

0.1 0.2 0.4 0.8

r

2a-.s

sr S

-

r s

-

r

--

s

1 r s

-

2 r

_

10

5

s

pH 7

0 1 5 1500.10.61 5 V

20

rss .s * i # * r

s

40 V* ..

am

4-

Aodk

:0 4m

.-

4.0 *

FIG. 1. PBP patterns of methicillin-sensitive strain 27 (s) and methicillin-resistant strain 27R (r) of S. aureus at pH 7.0. Concentrations of [3H]benzylpenicillin range from 0.1 to 20 p.g/ml. PBPs 1, 2, 2a, 3, and 4 are labeled.

A

A

FIG. 2. PBP pattern of methicillin-resistant strain Col in a competition experiment with various concentrations of methicillin (0 to 5 mg/ml). [3H]benzylpenicillin (10 ,ug/ml) was added after exposure to methicillin for 10 min (sequential competition). Lane 0 is a control lane in which only [3H]benzylpenicillin at 10 p.g/ml is added. Symbols: A, Concentration of methicillin at which PBP-2 disappeared; A, concentration at which PBP-2a disappeared.

Downloaded from jb.asm.org by on March 19, 2009

incubated at 37°C for another 30 min. After the addition of 25 ,u of sample dilution buffer (3), the entire sample was boiled again for 2 min. In one type of experiment, the methicillin and [3H]benzylpenicillin (at a constant concentration of 10 ,ug/ml) were added simultaneously to the bacteria, and incubation was for 10 min. In a second type of experiment (sequential competition) the bacterial sample was preincubated with the methicillin for 10 min, followed by the addition of [3H]benzylpenicillin (10 jig/ml) and an additional 10-min incubation. In both methods the samples were boiled, and the next procedure was followed as above. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (11) was done with a 10% acrylamide-0.13% bis-acrylamide concentration in the running gel measuring 13 by 18 cm. The stacking gel was composed of 5% acrylamide and 0.07% bisacrylamide. Samples were added to the gel lanes and run through the stacking gel at 75 mV and through the running gel at 120 mV. The samples were run for 45 to 60 min after the leading edge had passed the bottom of the running gel (total time through the running gel, ca. 4 h). The gels were stained with Coomassie brilliant blue (6) and destained overnight with methanol-acetic acid-water (30:5:65%). The gels were processed by first removing the water by two 30-min soaks in dimethyl sulfoxide, followed by a 2-h soak in a 20% (wt/vol) solution of the scintillator PPO (2,5-diphenyloxazole) in dimethyl sulfoxide. The gels were then rehydrated in 1% glycerol solution in water for 45 to 60 min, dried thoroughly, and exposed to a Kodak XOMAT XAR-2 film for 3 to 14 days at -70°C (1). Assay for penicillinase. The presence of penicillinase was assayed with Nitrocefin (Glaxo Research Ltd., Greenford, Middlesex, England) (13). RESULTS Figure 1 shows the PBPs of the methicillin-sensitive strain 27 and the isogenic methicillin-resistant transductant strain

27R, both grown at pH 7.0. Exposure of the preparations to a range of [3H]benzylpenicillin concentrations (0.01 to 1.0 P,g/ ml) revealed a rather similar band pattern. However, exposure to higher concentrations of the antibiotic (2 to 20 ,ug/ml) resulted in the appearance of an extra radioactive band (PBP-2a) in the methicillin-resistant strain. This PBP-2a ran slightly ahead of PBP-2 with an approximate molecular weight of 78,000. No PBP with similar mobility was observed in either strain 27 or in a second penicillin-sensitive control strain 209-P (data not shown). The detection of PBP2a required exposure to a minimum concentration of 2 to 5 p.g of penicillin per ml. This is close to the MIC for the resistant strain 27R (Table 1). An extra PBP, indistinguishable from PBP-2a, was also detectable in strain Col, a naturally occurring clinical isolate (Fig. 2). The low affinity of PBP-2a to methicillin could be demonstrated by competition experiments at pH 7.0 (Fig. 2). Exposure of the methicillin-resistant Col strain to various concentrations of methicillin 10 min before the addition of 10 ,pg of [3H]benzylpenicillin per ml resulted in a titration profile in which the familiar PBPs (1 and 3) were completely saturated by 50 ,ug of methicillin per ml; PBP-2 required 100 to 600 ,ug of methicillin per ml to achieve similar saturation. However, binding of PBP-2a required still higher concentrations, in the range of 1,000 ,ug of methicillin per ml. Virtually identical results were obtained with the methicillin-resistant transductant strain 27R (data not shown). The low affinity of PBP-2a for methicillin could also be demonstrated in simultaneous competition experiments (Fig. 3) in which resistant strain 27R was exposed simultaneously to methicillin (at various concentrations) plus [3H]benzylpenicillin. In this case, saturation of PBP-2a occurred at or above the MIC (625 ,ug/ml) of the organism for the antibiotic. A comparison of Fig. 3 with Fig. 2 indicates that in the simultaneous competition experiment, substantially higher concentrations of methicillin (up to 600 ,ug/ml) were also needed for the saturation of PBPs 1, 2, and 3 than were

515

METHICILLIN-RESISTANT S. AUREUS: PBPs AT pH 7 AND 5.2

VOL. 158, 1984

DISCUSSION Previous work with the PBPs of methicillin-resistant staphylococcal strains has resulted in the description of a variety of PBP alterations associated with resistance to betalactam antibiotics. Brown and Reynolds compared the PBPs of a resistant and a sensitive strain (2). They reported both a normal PBP pattern and a high beta-lactam affinity in the PBPs of both strains, with the exception of PBP-3 which, in the resistant strain, appeared to have lower affinity for penicillin and other beta-lactam antibiotics. In addition, under growth conditions in which the resistant strain could express its methicillin resistance (growth at 30°C), a protein with an electrophoretic mobility similar to PBP-3 appeared, and a corresponding heavy band appeared on the Coomassie brilliant blue-stained gel. These authors suggested that in

0.1

0.2

OA

r s

r s

r s

r

s

2-

ak

af

3-

5

r s

1-

I

s

r

_

___ ot a

O-

aO

r

`

..

~w3 O

FIG. 4. PBP profiles of methicillin-sensitive strain 27 (s) and methicillin-resistant strain 27R (r) in a [3H]benzylpenicillin titration experiment (1 to 20 p.g/ml). The titration was performed in growing bacterial cultures at pH 5.2. PBPs 1, 2, and 3 are labeled. PBP-4 is not seen well for either strain. PBP-2a is absent for both strains.

their strains, resistance may result from either an increased amount of PBP-3 or the presence of a new low affinity PBP with a molecular weight identical to PBP-3. Georgopapadakou et al. (8) reported that PBP-3 may be missing (or present with a reduced binding affinity) in a cephradine-resistant clinical isolate of S. aureus. This isolate also appeared to have an increased concentration of a low affinity PBP-2, as well as a satellite band (PBP-2') with an approximate molecular weight of 78,000. These authors suggested that antibiotic resistance was related to poor binding to PBP-2. PBP-2' did not show decreased affinity for cephradine and other relevant cephalosporins. This strain had relatively low MICs for methicillin and many other betalactam antibiotics and thus may not be directly comparable to methicillin-resistant strains studied by others (2, 9, 10). In addition, the strains compared were not isogenic. Hayes et al. (10), studying another methicillin-resistant clinical strain, showed that PBP-3 had relatively low affinities for beta-lactam antibiotics which were consistent with the respective MICs. As a comparison, nonisogenic methicillin-sensitive strain H was used. Utsui et al. (Y. Utsui, M. Tajima, R. Sekiguchi, E. Suzuki, and T. Yokota. Abstr. Int. Congr. Chemother. 13th, Vienna, Austria, abstr. no. 2.11/3-2, 1983) reported that cephemresistant clinical isolates of S. aureus possessed a new 78,000-dalton protein with low affinities to benzylpenicillin and some other beta-lactam antibiotics. Spontaneous revertants were found to have lost the 78,000 PBP. However, this 78,000 PBP had a slower mobility than PBP-2 in contrast to the PBP-2' (8) observed in cephradine-resistant S. aureus, which migrated ahead of PBP-2. The relationship of the cephem-resistant organisms to the methicillin-resistant staphylococci is not clear.

PH15.2

*PH 7

I ug'" --...... --. - IMQ 0.2 0.6 1 0 Oi 1 5 10 20 50 01.

s

2

1

0.8 r

ID W0 QS1 5 O 1

KIO 50al"5

5

......

5

FIG. 3. PBP pattern for strain 27R in a competition experiment with various concentrations of methicillin (O to 5 mg/ml) for 10 min. [3H]benzylpenicillin (10 ,ug/ml) was added simultaneously with the methicillin (simultaneous competition).

77'0* ~ ~

~

~

~

-

-

.

~

--2

FIG. 5. PBP pattern of methicillin-resistant strain Col in a competition experiment with various concentrations of methicillin (0 to 5 mg/ml) at both pH 7.0 and 5.2. [3H]benzylpenicillin (10 p.g/ml) was added 10 min after exposure to methicillin (sequential competition). PBPs 1, 2, 2a, and 3 are labeled. Lanes 0 are controls to which only [3H]benzylpenicillin was added.

Downloaded from jb.asm.org by on March 19, 2009

needed in the sequential competition experiments (Fig. 2 and data not shown). Similar results have already been described (9). The visualization of PBP-2a made it necessary to use high concentrations of radioactive penicillin which, in turn, has led to the appearance of increased background radioactivity. Similar situations have also been noted with other bacterial species (18). The extra bands detectable at high penicillin concentrations represent, presumably, nonspecific (nonenzymatic) binding of the antibiotic. This is supported by the observation that in the methicillin competition experiments (see, e.g., Fig. 2) the intensity of background bands did not decrease in contrast to the behavior of the true PBPs. Figure 4 shows the PBP profiles of methicillin-sensitive strain 27 and its isogenic transductant 27R grown at pH 5.2. Examination of the fluorogram indicates the similarity of PBP patterns and the absence of PBP-2a from among the PBPs of strain 27R. By comparing Fig. 1 with Fig. 4, one can see the extra binding protein (PBP-2a) in the methicillinresistant transductant grown at pH 7.0. A culture of the resistant clinical isolate (Col) grown at two pH values also yielded a similar PBP pattern: PBP-2a was only demonstrable in the cultures grown at pH 7.0 (Fig. 5, lane 0). The absence of PBP-2a was not due to slower penicillin binding since this protein was not detectable even upon prolonged (1 h) exposure to penicillin (data not shown). The methicillin competition experiments were run with the same clinical isolate (Col) grown at two pH values. PBP-2a (demonstrable in the pH 7 cultures) had a low affinity for methicillin, since saturation of this binding protein required more than 1 mg of methicillin per ml (Fig. 5). At pH 7, all the other high-molecular-weight PBPs (1, 2, and 3) were saturated below 100 xg of methicillin per ml. In the cultures grown at pH 5.2, PBP-2a was not detectable, and the other three high-molecular-weight PBPs were saturated below a concentration of 100 .g of methicillin per ml.

516

HARTMAN AND TOMASZ

LITERATURE CITED 1. Bonner, W. M., and R. A. Laskey. 1974. A film detection method for tritium-labeled proteins and nucleic acids in poly-

acrylamide gels. Eur. J. Biochem. 46:83-88. 2. Brown, D. F. J., and P. E. Reynolds. 1980. Intrinsic resistance to ,B-lactam antibiotics in Staphylococcus aureus. FEBS Lett. 122:275-278. 3. Buchanan, C. E., and J. L. Strominger. 1976. Altered penicillinbinding components in penicillin-resistant mutants of Bacillus subtilis. Proc. Natl. Acad. Sci. U.S.A. 73:1816-1820. 4. Centers for Disease Control. 1981. Methicillin-resistant Staphylococcus aureus-United States. Morbid. Mortal. Weekly Rep. 30:557-559. 5. Dougherty, T. J., A. E. Koller, and A. Tomasz. 1980. Penicillinbinding proteins of penicillin-sensitive and intrinsically resistant Neisseria gonorrhoeae. Antimicrob. Agents Chemother. 18:730-737. 6. Fairbanks, G. T., T. L. Steck, and D. H. F. Wallach. 1971. Electrophoretic analysis of the major polypeptides of the human erythrocyte membrane. Biochemistry 10:2606-2617. 7. Fontana, R., R. Cerini, P. Longoni, A. Grossato, and P. Canepari. 1983. Identification of a streptococcal penicillin-binding protein that reacts very slowly with penicillin. J. Bacteriol. 155:1343-1350. 8. Georgopapadakou, N. H., S. A. Smith, and D. P. Bonner. 1982. Penicillin-binding proteins in a Staphylococcus aureus strain resistant to specific P-lactam antibiotics. Antimicrob. Agents Chemother. 22:172-175. 9. Hartman, B., and A. Tomasz. 1981. Altered penicillin-binding proteins in methicillin-resistant strains of Staphylococcus aureus. Antimicrob. Agents Chemother. 19:726-735. 10. Hayes, M. V., N. A. C. Curtis, A. W. Wyke, and J. B. Ward. 1981. Decreased affinity of a penicillin binding protein for ,Blactam antibiotics in a clinical isolate of Staphylococcus aureus resistant to methicillin. FEMS Microbiol. Lett. 10:119-122. 11. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227:680-685. 12. Mirelman, D., Y. Nuchamowitz, and E. Rubinstein. 1981. Insensitivity of peptidoglycan biosynthetic reactions to P-lactam antibiotics in a clinical isolate of Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 19:687-695. 13. O'Callaghan, C. H., A. Morris, S. M. Kirby, and A. H. Shingler. 1972. Novel method for detection of P-lactamases by using a chromogenic cephalosporin substrate. Antimicrob. Agents Chemother. 1:283-288. 14. Rosegay, A. 1981. High specific activity (Phenyl-3H) benzylpenicillin N-ethylpiperidine salt. J. Labelled Compd Radiopharm. 18:1337-1340. 15. Sabath, L. D. 1977. Chemical and physical factors influencing methicillin resistance of Staphylococcus aureus and Staphylococcus epidermidis. J. Antimicrob. Chemother. 3(Suppl. C):4751. 16. Sabath, L. D., S. J. Wallace, and D. A. Gerstein. 1972. Suppression of intrinsic resistance to methicillin and other penicillins in Staphylococcus aureus. Antimicrob. Agents Chemother. 2:350355. 17. Saravolatz, L. D., D. J. Pohlod, and L. M. Arking. 1982. Community-acquired methicillin-resistant Staphylococcus aureus infections: a new source of nosocomial outbreaks. Ann. Intern. Med. 97:325-329. 18. Schwarz, U., K. Seeger, F. Wengenmayer, and H. Strecker. 1981. Penicillin binding proteins of Escherichia coli identified with a 1251-derivative of ampicillin. FEMS Microbiol. Lett. 10:107-109. 19. Spratt, B. G. 1978. Escherichia coli resistance to 3-lactam antibiotics through a decrease in the affinity of a target for lethality. Nature (London) 274:713-715. 20. Thompson, R. L., I. Cabezudo, and R. P. Wenzel. 1982. Epidemiology of nosocomial infections caused by methicillin-resistant Staphylococcus aureus. Ann. Intern. Med. 97:309-317. 21. Zighelboim, S., and A. Tomasz. 1980. Penicillin-binding proteins of multiply antibiotic-resistant South African strains of Streptococcus pneumoniae. Antimicrob. Agents Chemother. 17:434442.

Downloaded from jb.asm.org by on March 19, 2009

We compared a methicillin-sensitive strain of S. aureus to its isogenic methicillin-resistant transductant. In contrast to previous studies, we used labeling of PBPs by live growing bacteria. This method should allow a more appropriate comparison of PBP saturation with the MIC than a method with membrane preparations. None of the strains had detectable P-lactamase activity. The comparison of the penicillin titration profiles of the two strains indicated minor differences in the quantity (or affinities?) of PBPs 1, 3, and 4 (more in the sensitive strain) and a possible minor difference in the electrophoretic mobilities of PBP-2 (slightly slower mobility in the sensitive strain). However, the major difference between the PBP patterns of the two strains was only detectable at high concentrations of penicillin. At these higher concentrations, only the resistant strains showed an extra PBP (PBP-2a) running between PBP-2 and -3 at pH 7.0. An indication of the existence of PBP-2a is seen in the fluorograms published in our earlier communication (see Fig. 8, in reference 9). However, the complete separation of this PBP from PBP-2 required the improved technique used in the present work. This PBP-2a was also detected in a methicillin-resistant clinical isolate (strain Col) which had no beta-lactamase activity. The competition experiments with methicillin indicated that PBP-2a had a low affinity for methicillin. The fact that the concentrations of both benzylpenicillin and methicillin needed to achieve saturation of this PBP were close to the biologically effective concentrations (MICs) suggests that the beta-lactam resistance in these strains may be related to the acquisition of this low-affinity protein. This additional PBP may allow continued cell wall synthesis in the presence of antibiotics in the medium at elevated concentrations such that the other PBPs would be inactivated. This suggestion is similar to the one already proposed by Fontana et al. (7) for the possible function of low-affinity binding protein 5 detectable in the beta-lactam-resistant strains of Streptococcus faecalis and Streptococcus faecium. The inability of methicillin-resistant staphylococci to express their antibiotic resistance when grown at low pH (5.2) is a general property of all resistant isolates. In an earlier report (9) we noted that there were no significant changes in the molecular weight and drug affinities of the three highmolecular-weight PBPs when the cultures were grown at either pH 7.0 or 5.2. We report now the resolution and separation of an additional PBP (PBP-2a) present in cultures grown at pH 7.0 but not in cultures grown at pH 5.2. This correlates well with the expression of resistance at the two pH values. One would expect that the bacterial cell wall synthesis catalyzed in growing bacteria by a normal complement of PBPs at pH 7.0 would be slowed or altered in the presence of high concentrations (sub-MIC) of beta-lactams, since only PBP-2a would be available for the catalysis of the cell wall synthetic reactions. It is likely that cell wall synthesized under these conditions would have abnormal structure. We are presently investigating these possibilities. Further experiments will also be needed to determine whether the disappearance of PBP-2a at pH 5.2 is due to its instability, nonfunctioning state, or inhibition of the biosynthesis of this protein in low pH cultures.

J. BACTERIOL.