Penicillin-Resistant and Penicillin-Tolerant Mutants of Group

ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, July 1982, p. 128-136 0066-4804/82/070128-09$02.OO/O Vol. 22, No. 1 Penicillin-Resistant and Penicillin-Tolerant Mutants of Group A Streptococci LAURENT GUTMANN AND ALEXANDER TOMASZ* The Rockefeller University, New York, New York 10021 Received 19 February 1982/Accepted 28 April 1982 Penicillin-resistant and penicillin-tolerant mutants have been isolated from group A streptococci mutagenized by ethyl methane sulfonate. The resistant mutants had an elevated minimal growth inhibitory concentration for benzylpenicillin (minimal inhibitory concentration, 0.2,ug/ml, as compared with the minimal inhibitory concentration of 0.006 jtg/ml in the penicillin-susceptible parent strain); they also had an abnormal cellular morphology and showed altered penicillinbinding proteins. Penicillin-tolerant mutants were killed more slowly than were the parental cells during treatment with penicillin; they had virtually unchanged minimal inhibitory concentration values for penicillin and normal cellular morphology and penicillin-binding proteins. Group A streptococci are causative agents of a variety of human diseases, and penicillin has been widely used in clinical medicine as a chemotherapy of choice. This antibiotic is known to cause rapid loss of viability in growing cultures of group A streptococci without apparent lysis of the cells (12). Although resistant mutants have been isolated in the laboratory, there has been no report of a shift toward a higher level of penicillin resistance in natural isolates, despite the extensive use of penicillin for prophylaxis in acute and chronic forms of the disease. Penicillin resistance has never been observed during the routine testing of the antibiotic susceptibilities of group A streptococci isolated in cases of treatment failures (R. Facklam, Centers for Disease Control, Atlanta, Ga., personal communication). In contrast, the emergence of penicillinresistant forms has been frequently reported among a-hemolytic streptococci of the oral flora of patients on prolonged penicillin therapy (20). Penicillin-tolerant isolates, i.e., group A streptococci with increased penicillin minimal bactericidal concentration values relative to the penicillin minimal inhibitory concentration (MIC) values have been observed in clinical specimens (1). Up to the late 1960s, resistance to non-betalactam antibiotics was rare in group A streptococci. Since the early 1970s, however, large increases in the frequency of resistance of clinical isolates to sulfonamides, tetracycline, chloramphenicol, and erythromycin have been reported, and several hospital surveys have indicated that such resistant bacteria may make up a substantial fraction of current clinical isolates. In addition, mechanisms for the cell-tocell transfer of resistance genes have also been described (13, 24). The emergence of resistance in group A streptococci is reminiscent of the situation described for pneumococci in the late 1970s, when multiply antibiotic-resistant organisms carrying high levels of penicillin resistance suddenly appeared in hospital isolates in several locations (14). These penicillin-resistant pneumococcal strains were shown to contain altered penicillin-binding proteins (PBPs) (25). In the present communication, we describe the laboratory isolation and preliminary characterization of penicillin-resistant and penicillintolerant group A streptococcal mutants. It is hoped that future studies with these mutants will provide useful information about the mechanism of action of penicillin in these bacteria. MATERIALS AND METHODS Bacterial strains and growth conditions. Streptococcus pyogenes T4/56 (obtained from M. McCarty, The Rockefeller University, New York, N.Y.) was grown in Todd-Hewitt broth (THB) (Difco Laboratories, Detroit, Mich.) in 10-ml batch cultures in 18-mm culture tubes at 37 C in air without shaking. Growth was monitored with a Coleman nephelometer. Isolation of penicillin-resistant mutants. An exponentially growing culture of S. pyogenes T4/56 in THB at a cell concentration of about 7 x 107 colony-forming units (CFU) per ml was treated with ethyl methane sulfonate (150 1±1V10 ml of culture) at 37 C for 2 h (16). This treatment caused the loss of about 90% of the CFU. After removal of the mutagen, the surviving bacteria were grown in THB, and 0.1-mi portions of this culture (containing about 8 x 106 CFU) were used to inoculate test tubes containing 10 ml of THB and a series of concentrations of penicillin ranging from 0.003 to 0.2,ug/ml. The cultures were incubated at 37 C for 48 h; 0.1 ml of bacteria from the tube with the 128

VOL. 22, 1982 highest concentration of penicillin which permitted growth was used as an inoculum for a second set of culture tubes containing penicillin, as described above (17). Bacteria growing in the medium with 0.2 p.g of penicillin per ml were plated on brain heart infusion agar plates containing 5% blood and penicillin (0.1 pg/ ml). Resistant mutants were purified by two rounds of single-colony isolation. Stocks were stored in THB medium with 10% glycerol at -70 C. A single mutant isolate, P23, was used for detailed study. Stability of the penicillin-resistant mutant P23. Five colonies (P423, P623, P723, P923, and P1023) derived from the resistant mutant P23 were subcultured 10 times (100-fold dilution each time) in penicillin-free medium, and the plating efficiency of the isolates was compared on penicillin-free and penicillin-containing agar plates. The test was done in the following way. One hundred to 300 CFU were plated on penicillin-free blood agar master plates. After 18 h at 37 C, a replica plating was done onto a series of blood agar plates containing increasing concentrations of penicillin. At 0.1,ug of penicillin per ml, which was the highest concentration at which P23 was still able to grow, the plating efficiency was 90 to 100%, except for P923, which had a low plating efficiency (about 2%) at this concentration of penicillin but was fully able to grow on plates with 0.05 jig of penicillin per ml. Tolerant mutants. Tolerant mutants were obtained from a mutagenized culture as described above. A 0.1- ml amount of the mutagenized culture was inoculated into 10 ml of THB and incubated until the optical density reached 200 nephelometric units (approximately 5 x 107 cells per ml), at which time 0.1 p.g of penicillin per ml was added (about 15 times the MIC). After 6 h of incubation, the cells were washed two times with sterile THB (centrifugation) and suspended in 10 ml of antibiotic-free THB. One milliliter of this suspension was inoculated into 9 ml of THB and incubated overnight. The penicillin treatment (0.1,g/ ml for 6 h) was then repeated six times to allow the enrichment of penicillin-tolerant cells. Finally, 0.1 ml of such a culture (approximately 5 x 106 CFU) was spread onto blood agar plates containing 0.2 p.g of pencillin per ml and incubated for 12 h at 37 C. Surviving bacteria were replica plated with sterile velvet onto blood agar plates without penicillin and incubated at 37 C for 40 h. About 40 colonies were isolated on each plate by this method. Each clone used in this work was retested for the homogeneity of its tolerant response to penicillin in the following manner. About 100 to 300 CFU were plated on blood agar master plates, and after 12 h of incubation at 37 C, the colonies were replica plated onto blood agar plates containing 0.2,ug of penicillin per ml. These latter plates were incubated at 37 C for 12 h, at which time a second replica plating was done onto blood agar plates free of penicillin. Virtually all of the colonies survived this penicillin challenge, as evidenced by the recovery of 95 to 100% of the colonies of the master plate. The same treatment resulted in the loss of 100% of the parental wild-type streptococci (no survivors among 200 to 300 colonies). Two tolerant mutants, Tol23 and TolS7, were chosen for detailed study. Testing of antibiotic susceptibility. The MIC was determined by the tube dilution method (21) in THB. With an inoculum of 104 to 105 CFU/ml and an incubation time of 20 h, the viable titer of cultures was PENICILLIN-RESISTANT GROUP A STREPTOCOCCI 129 TABLE 1. MIC values for various antibiotics in the penicillin-sensitive (wild-type), penicillin-resistant (P23), and penicillin-tolerant (Tol23) strains of group A streptococci MIC (,ug/ml) of following Antibiotic strain: Wild type P23 Tol23 Penicillin G' 0.006 0.2 0.006 Oxacillin 0.03 1 0.03 Ampicillin 0.007 0.25 0.007 Cefoxitin 1 2 0.5 Cephaloridine 0.007 0.03 NDb Cefsulodin 4 4 ND Cefotaxime 0.015 0.5 ND Thienamycin (MK 0787) 0.0035 0.06 0.0035 Sch 29482 0.03 0.5 0.03 Vancomycin 0.5 0.5 0.5 Cycloserine 180 180 ND Gentamicin 2 0.5 4 Streptomycin 8 4 16 Amikacin 8 2 32 Tetracycline 0.12 ND 0.12 Erythromycin 0.015 0.03 0.03 Rifampin 0.06 0.03 0.12 Aerosporin 16 8 ND a In another penicillin-tolerant mutant (Tol57), the MIC for penicillin was 0.01 p.g/ml. b ND, Not determined. determined by the standard procedure for counting colonies on blood agar plates. Variation in the inoculum size (between 103 and 105 CFU/ml) did not change the MIC values. Organisms were diluted in saline; 0.1- ml volumes of each sample were spread onto plates and incubated for 36 h at 37 C. Osmotic protection of penicillin-treated cells. Bacteria grown and treated with penicillin in THB were plated on blood agar plates supplemented with hypertonic sucrose and sodium chloride (brain heart infusion broth; 1.2% agar-0.6 M sucrose-0.25 M NaCI). Alternatively, the bacteria were grown and treated with penicillin in liquid THB medium supplemented with sucrose and sodium chloride at the above concentrations and then plated on agar with or without osmotic supplements. Leakage of prelabeled protein and RNA from group A streptococci. An exponential-phase culture of bacteria in THB supplemented with uracil (5,ug/ml) was divided into two 12-ml samples which received either 3,uCi of [3H]uracil or 18 pci of [35S]methionine per ml. Further incubation at 37 C was continued for four generations. The cells were harvested (12,000 x g, 5 min) and washed three times with cold THB. Each sample was suspended in 25 ml of THB and allowed to grow for 30 min, at which time half of the culture samples received penicillin (0.15 p.g/ml). Further incubation was done at 37 C. At intervals, 500-pIl samples were centrifuged at 12,000 x g for 10 min. Radioactive material released into the supernatant was assayed in the following way. Two hundred fifty microliters of the supernatant received 2 ml of cold 10% trichloroacetic acid and 25 VI of 4% albumin solution (Armour fraction V). The precipitates were collected on glass fiber (GFA) filters and radioactively determined in a mark II

130 GUTMANN AND TOMASZ ANTIMICROB. AGENTS CHEMOTHER. z I-I 0 C, HOURS FIG. 1. Inhibition of growth of wild-type and penicillin-resistant streptococci by penicillin. Exponentially growing cultures of wild-type (WT) and penicillin-resistant (P23) cells were exposed to several concentrations of penicillin (added at the time indicated by the arrows) ranging from 0.01 to 2.0 jg/ml (see numbers on graphs). Control cultures (C) received no antibiotic. spectrometer (Nuclear-Chicago Corp., Des Plaines, Ill.). In vivo labeling of PBPs. One-milliliter samples of exponentially growing bacteria (about 5 x 107 CFU/ ml) were diluted to identical optical density values and exposed to a series of concentrations of [3H]penicillin for 10 min at 37 C. Nonradioactive penicillin (500 pg) was added, and the bacteria were recovered by centrifugation (12,000 x g at 4 C for 5 min). The cells were suspended in 40 RI of lysis buffer (50 mm sodium phosphate buffer [ph 6.1] containing 5 mm EDTA) in the presence of an excess of nonradioactive penicillin. A lytic enzyme (C-phage associated lysin) prepared by the method of Fischetti et al. (6) was added, and the samples were incubated at 37 C for 10 min. Wild-type and mutant cultures appeared to undergo complete lysis under these conditions, as judged by both the complete clearing of suspensions and the pattern of protein bands visualized by Coomassie blue staining. After the addition of a 25-,ul sample of dilution buffer (3), the samples were boiled for 2 min and applied to polyacrylamide slab gels. Polyacrylamide slab gel I I FIG. 2. Electron micrographs of the penicillin-sensitive (wild-type) and penicillin-resistant (P23) streptococci. Scanning electron micrographs of the wild-type (A) and mutant (B) cells and thin section of the mutant bacteria (C) are shown. Bars: (A) and (B), 2 Rim; (C), 0.5 pum.

VOL. 22, 1982 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~ - FIG. 3. PBPs of penicillin-tolerant and penicillinresistant group A streptococcal mutants. Exponentially growing cultures of wild-type (WT) (strain T4/56) streptococci and the penicillin-tolerant (Tol57 [T57]) and penicillin-resistant (P23) strains were exposed to the same series of concentrations of [3H]penicillin, and the PBPs were identified by fluorography as described in the text. The penicillin concentrations (in micrograms per milliliter) were as follows: lane a, 7.5; lane b, 2; lane c, 1; lane d, 0.2; lane e, 0.01; lane f, 0.05; lane g, 0.01; lane h, 0.005; lane i, 0.001. electrophoresis and detection of PBPs were carried out as previously described (8). Electron microscopy. Bacteria were grown in THB. Some samples were fixed in 2% glutaraldehyde (in 0.05 M potassium phosphate buffer [ph 7.0]) and processed for scanning electron microscopy. Other samples of the penicillin-resistant bacteria were fixed, first in glutaraldehyde (2%) and then in 0.5% OS04; prestained with uranyl acetate; and subsequently processed for transmission electron microscopy (19, 23). Thin sections were examined with an EM4 electron microscope (Philips Electronic Instruments, Inc., Mahwah, N.J.) at an accelerating voltage of 110 kv. Antibiotics and reagents. [3H]benzylpenicillin (25 Ci/ mmol) (A. Rosegay, J. Labelled Compd. Radiopharm., in press) was obtained from Merck Sharp & Dohme, West Point, Pa. Other nonlabeled compounds included: benzylpenicillin, cephaloridine, and vancomycin (Eli Lilly & Co., Indianapolis, Ind.); cefoxitin PENICILLIN-RESISTANT GROUP A STREPTOCOCCI 131 (MK 0787; Merck Sharp & Dohme); cefotaxime (Hoechst-Roussel Pharmaceuticals Inc., Somerville, N.J.); oxacillin, ampicillin, and amikacin (Bristol Laboratories, Syracuse, N.Y.); gentamicin (Schering Corp., Bloomfield, N.J.); streptomycin, erythromycin, tetracycline, rifampin, and cycloserine (Sigma Chemical Co., St. Louis, Mo.); aerosporin (Burroughs Wellcome Co., Research Triangle Park, N.C.); Sch 29482 (Schering Corp.); and cefsulodin (SCE-129; Takeda Chemical Industries Ltd., Osaka, Japan). Ethyl methane sulfonate was purchased from Sigma Chemical Co., [35S]methionine (1,350 Ci/mmol) was obtained from Amersham Corp., Arlington Heights, Ill., and [3H]uracil (23.3 Ci/mmol) was obtained from New England Nuclear Corp., Boston, Mass. RESULTS The penicillin-resistant bacterium had increased MIC values for a number of structurally diverse beta-lactam antibiotics (Table 1). However, the degree of change in the MIC values varied greatly from one beta-lactam to another. For instance, whereas the MIC values for benzylpenicillin and oxacillin increased about 30- fold, there was relatively little increase (two- to fourfold) in the MIC values for cefoxitin or cephaloridine and none for cefsulodin (Table 1). Susceptibilities to several non-beta-lactam antibiotics remained virtually unchanged. There was a somewhat increased susceptibility to aminoglycosides in mutant P23. Penicillin-resistant mutants. Inhibition of growth of the penicillin-resistant mutant P23 (benzylpenicillin MIC, 0.2 Kg/ml) required substantially higher concentrations of penicillin than did inhibition of the parental penicillinsusceptible strain T4/56 (benzylpenicillin MIC, 0.006 p.g/ml) (Fig. 1). Strain P23 also grew more slowly than strain T4/56 in THB (average doubling times, 108 min for P23 and 35 min for T4/ 56). Addition of penicillin resulted in rapid loss of viability in both strains (data not shown), although the bactericidal effect of penicillin was somewhat slower in P23, perhaps owing to the slower growth rate. Cells of strain P23 grown in THB had an aberrant morphology (Fig. 2A and B). The abnormal morphology of several subclones was shown to be retained during at least 10 subcultures (involving about 100 cell generations) in penicillin-free THB, and such cultures had the same plating efficiencies on drug-free and on penicillin-containing (0.1,ug/ml) blood agar plates. Exponentially growing cultures of T4/56 and P23 were briefly exposed to various concentrations of radioactive penicillin, and the labeling patterns of the PBPs were compared (Fig. 3). Quantitative scanning of the labeled PBPs revealed several alterations in PBP 2 of the resistant mutant, suggesting both lower cellular con-

132 GUTMANN AND TOMASZ en zwa ANTIMICROB. AGENTS CHEMOTHER. A a z 4m z "a C] z 4 600 500 L 400[ 300 0.2 0.1 0.05 pg/mi 0.01 0.005 0.001 B 2001.._._a. I 1001. ~~ ;-- ~~ 10 2 1 0.2 0.1 0.05 0.01 0.005 0.001 pg/ml FIG. 4. Quantitation of the radioactive penicillin bound to PBPs 2 (A) and 3 (B) of wild-type and mutant streptococci. The fluorograms shown in Fig. 3 were quantitated by scanning densitometry as described in the text (15). Curves for Tol23 and resistant strain P723 (clone from P23 which was subcultured for 10 passages in penicillin-free medium) are derived from scans of other gels done in the same experiment (not shown). Symbols: *, wild type; A, Tol57; U, Tol23; [, P23; A, P723. centration and lower penicillin affinity. The resistant bacteria also appeared to have an altered PBP 3; this protein seemed to have lower antibiotic affinity and was present at a higher concentration than in the susceptible bacteria (Fig. 3 and 4). Penicillin-tolerant mutants of group A streptococci. Several attempts were made to detect cell structural defects in group A streptococci killed by penicillin. Penicillin treatment that caused loss of over 99% of the CFU caused no detectable loss of protein or nucleic acid material from bacteria (loss of less than 1% of the total incorporated label precipitable by cold trichloroacetic acid) prelabeled with the appropriate radioactive isotopes. In addition, an examination of electron microscopic sections of penicillin-treated bacteria revealed no clear sign of structural damage (Fig. 5). On the other hand, a parallel assay of viability of penicillin-treated bacteria on normal blood agar plates and blood agar plates supplemented by high concentrations of sucrose and sodium chloride indicated that a large portion of the cells could be salvaged as viable colony formers on the osmotically supplemented plates (Fig. 6). This finding suggested that, similar to other species of bacteria, group A streptococci may lose viability by a process involving structural damage to the cell wall. If this were the case, then it might be possible to find mutants of group A streptococci specifically blocked in the process(es) involved with the viability loss. We succeeded in isolating such penicillin-tolerant mutants. Figure 7 shows a comparison of the responses of the parental and two mutant cultures (Tol23 and Tol57) to penicillin treatment. All three cultures grew with comparable rates and showed inhibition of growth by penicillin at the same concentrations and by comparable kinetics. The tolerant mutants had normal morphology. On the other hand, penicillin-induced loss of viabili-

<.. -- - - - - -C VOL. 22, 1982 PENICILLIN-RESISTANT GROUP A STREPTOCOCCI 133 B 0.5pm 0.5 pm FIG. 5. Morphology of penicillin-treated and control group A streptococci. Bacteria grown in THB were treated with penicillin (20x MIC) for 2 h as in the experiment described in Fig. 1. Control (A) and treated (B) cells were fixed with 2% glutaraldehyde added to the growth media. After a second fixation in OS04 (0.5%) and staining with uranyl acetate, the cells were put through a routine cytological procedure (19, 23). ty occurred only at greatly reduced rates in the ing the plates with penicillinase and a subsequent overnight incubation. No viable cells tolerant cultures. The tolerance of mutant cultures was not due could be recovered from similarly treated wildtype (T4/56) cells. Table 1 shows that the toler- to the presence of a subpopulation of slowgrowing (and, thus, phenotypically tolerant) ant mutants and wild-type streptococci had virtually identical MIC values for beta-lactam cells, since after exposure of several hundred tolerant cells on agar plates containing penicillin antibiotics. However, a modest increase in the (0.2,ug of penicillin per ml for 12 h), all bacteria MIC values of several aminoglycosides was noted. A preliminary examination of the PBPs of osm could be rescued as colony formers after flood-. * C bap - - - C bap - ~~; - - ------ - PEN osm 10 osm I. 10 0 6 10 m --------0------OP \ ~ ~ ~ ~~ ~ ~~ PEN \ osm PEN bap a PEN bap A B 2 3 0 HOURS 2 3 FIG. 6. Osmotic protection of group A streptococci treated with penicillin. (A) An exponential culture of wild-type group A streptococci (T4/56) was divided into two 12-ml samples. One received 0.12 p.g of penicillin (PEN), and the other served as a control (C). During further incubation at 37 C, samples were taken and, after appropriate dilutions, were spread on blood agar plates (bap) and osmotic plates (osm) (see text) at the same time. (B) An exponential culture of wild-type strain T4/56 was grown in liquid osmotic medium and divided into two 12-ml samples. One served as a control (C), and one received 0.12,ug of penicillin (PEN) per ml (see text). Further incubation was done at 37 C. Samples were taken at intervals, diluted in osmotic medium, and spread at the same time on blood agar plates (bap) and osmotic plates (osm).

134 GUTMANN AND TOMASZ ANTIMICROB. AGENTS CHEMOTHER. 1000 z 0 10 A - 0-0.0_2------ c -: - : i * : -0.02 a -- ---- - -_0.2 _ IO- - --- O_ 0.2 K 9.~~~~~~~.6 O C U-.H OL(2 8 'eq02 3WoT TQ02 FIG7.Efeto'eiilno 2 rwhadvaiiyo idtp W)adpncli-oeatmtn tan B D 1 2 4 16 S6 1 2 4 16 36 HOURS FIG. 7. Effect of penicillin on growth and viability of wild-type (WT) and penicillin-tolerant mutant strains (Tol57 and Tol23) of group A streptococci. Penicillin was added at different concentrations (0.02, 0.08, and 0.2,ug/ml) to exponential-phase cultures of T4/56 and its penicillin-tolerant mutants Tol57 and Tol23. (A) and (C), Effect on culture growth; (B) and (D), effect on cellular viability. Microscopic examination of the tolerant and wild-type (T4/56) cultures indicated no significant differences in degrees of chaining. Both kinds of cultures contained a distribution of chain sizes between 5 and 25 to 30 coccal units. (-----), Tolerant cultures; ( ), wild-type cells. Numbers on the graphs indicate the concentrations of penicillin used in micrograms per milliliter. C, Control cultures. tolerant mutants showed a normal complement of PBPs, with a somewhat lower concentration in one of the minor PBPs (PBP 3) (Fig. 3). DISCUSSION Although penicillin-resistant group A streptococci have not been reported among clinical isolates, mutants exhibiting elevated MIC values for penicillin have been constructed in the laboratory (5, 17). Some of these resistant isolates were shown to have a lower capacity for the cellular binding of radioactively labeled penicillin (5). The present report, in essence, confirms and extends these earlier observations: in penicillin-resistant mutants, PBPs 2 and 3, proteins responsible for the binding of over 60% of penicillin, were shown to have a decreased affinity for penicillin, and PBP 2 may also be present in lower cellular concentrations. Similar types of PBP alterations have been observed in both laboratory and clinical isolates of several other species of beta-lactam-resistant bacteria, and such PBP alterations may indeed be generally associated with intrinsic beta-lactam resistance (2-4, 7, 9, 10, 25). It is not known how many mutational events are involved with the penicillin-resistant and -tolerant phenotypes of the group A streptococci described in this report. The relative ease with which penicillin-resistant group A streptococci may be isolated in the laboratory makes one wonder why such bacteria have not appeared in natural isolates despite the extensive use of penicillin at low concentrations in various forms of group A streptococcal disease. One possible cause of this may be a selection against such bacteria in their natural environments, since an apparent decrease in the production of antiphagocytic M-protein has already been observed in the penicillin-resistant mutants of group A streptococci (18). Since PBPs are, presumably, enzymes involved with

VOL. 22, 1982 the biosynthesis of bacterial cell wall, penicillinresistant streptococci with their abnormal PBPs may produce sufficiently altered cell walls to hinder the synthesis or attachment or both of nonmurein wall polymers (M-protein) as well. The abnormal morphology of the penicillin-resistant mutants described here is consistent with this suggestion. Another phenomenon that may also contribute to the relative rarity (perhaps absence) of penicillin resistance among natural isolates of group A streptococci is the cooperative bactericidal effect of penicillins and leukocytes that has been observed during exposure of several bacterial species to sub-mics of penicillin. Studies are in progress to test these possibilities. The nature of the process that is altered in the tolerant mutants and that appears to be the ratelimiting step in the loss of viability of penicillintreated wild-type cells is not known at the present time. Penicillin treatment of group A streptococci stimulates the release of lipoteichoic acids and lipids into the surrounding medium (11). However, this appears to involve an active process rather than a correlate of viability loss, since streptococci tolerant to penicillin also exhibit penicillin-induced release of cell surface components (12). In other bacteria, penicillininduced lysis and killing can be correlated with the triggering of murein hydrolase activity and murein degradation (for review, see reference 22). It is conceivable that in group A streptococci also, the loss of bacterial viability involves triggering of a type of murein hydrolase activity that would introduce limited but irreparable nicks into the cell walls without causing complete collapse of cell structure. The observed rescue of penicillin-treated cells from loss of viability by plating on osmotically supplemented agar is consistent with such a mechanism. As an alternative, it has been suggested that in these bacteria, penicillin treatment may cause a membrane level defect, e.g., by activation of phospholipases (22). ACKNOWLEDGMENTS These investigations have been supported by Public Health Service grant Al 16170 from the National Institutes of Health and a fellowship to L.G. from the Ministere Francais des Affaires Etrangeres. We thank David Phillips (The Rockefeller University) for the scanning electron microscopy. LITERATURE CITED 1. Allen, J. L., and K. Sprunt. 1978. Discrepancy between minimum inhibitory and minimum bactericidal concentrations of penicillin for group A and group B hemolytic streptococci. J. Pediatr. 43:69-71. 2. Brown, D. F. J., and P. E. Reynolds. 1980. Intrinsic resistance to 1B lactam antibiotics in Staphylococcus aureus. FEBS Lett. 122:275-278. PENICILLIN-RESISTANT GROUP A STREPTOCOCCI 135 3. Buchanan, C. E., and J. L. Strominger. 1976. Altered penicillin binding components in penicillin-resistant mutants of Bacillus subtilis. Proc. NatI. Acad. Sci. U.S.A. 13:1816-1820. 4. Dougherty, T. J., A. E. Koller, and A. Tomasz. 1980. Penicillin-binding proteins of penicillin-susceptible and intrinsically resistant Neisseria gonorrhoeae. Antimicrob. Agents Chemother. 18:730-737. 5. Eagle, H. 1954. The binding of penicillin in relation to its cytotoxic action. II. The reactivity with penicillin of resistant variants of streptococci, pneumococci and staphylococci. J. Exp. Med. 100:103-115. 6. Fischetti, V. A., E. C. Gotschlich, and A. W. Bernheimer. 1971. Purification and physical properties of group C streptococcal phage associated with lysin. J. Exp. Med. 134:1105-1117. 7. Gilis, H. F., and P. E. Reynolds. 1979. Bacillus megaterium resistance to cloxacillin accompanied by a compensatory change in penicillin binding proteins. Nature (London) 280:167-168. 8. Gutmann, L., R. Williamson, and A. Tomasz. 1981. Physiological properties of penicillin-binding proteins in group A streptococci. Antimicrob. Agents Chemother. 19:872-880. 9. Hakenbeck, R., M. Tarpay, and A. Tomasz. 1980. Multiple changes of penicillin-binding proteins in penicillin-resistant clinical isolates of Streptococcuts pneumoniae. Antimicrob. Agents Chemother. 17:364-371. 10. Hayes, M. V., M. A. C. Curtis, A. W. Wide, and J. B. Ward. 1981. Decreased affinity of a PBP for P lactam antibiotics in a clinical isolate of Staphylococcus aureus resistant to methicillin. FEMS Microbiol. Lett. 10:119-122. 11. Horne, D., R. Hakenbeck, and A. Tomasz. 1977. Secretion of lipids induced by inhibition of peptidoglycan synthesis in streptococci. J. Bacteriol. 132:704-717. 12. Horne, D., and A. Tomasz. 1977. Tolerant response of Streptococcus sanguis to beta-lactams and other cell wall inhibitors. Antimicrob. Agents Chemother. 11:888-896. 13. Horodiceanu, T., L. Bougueleret, and G. Bieth. 1981. Conjugative transfer of multiple-antibiotic resistance markers in beta hemalytic group A, B, F and G streptococci in the absence of extrachromosomal deoxyribonucleic acid. Plasmid 5:127-137. 14. Koornhof, H. J., M. Jacobs, M. Isaacson, P. Appelbaum, B. Miller, C. M. Stevenson, I. Freeman, A. Naude, P. Botha, E. Glathaar, and J. Gilliland. 1978. Follow up on a multiple antibiotic resistant pneumococci. S. Afr. Morbid. Mortal. Weekly Rep. 27:1-7. 15. Laskey, R. A., and A. D. Mills. 1975. Quantitative film detection of 3H and "4C in polyacrylamide gels by fluorography. Eur. J. Biochem. 193:265-275. 16. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. 17. Rake, G., C. M. McKee, D. M. Hamre, and C. L. Houck. 1944. Studies on penicillin. J. Immunol. 48:271-283. 18. Rosendal, K. 1957. Investigation of penicillin resistant streptococci belonging to group A. Acta. Pathol. Microbiol. Scand. 39:181-188. 19. Simionescu, N., and M. Simionescu. 1976. Galloylglucoses of low molecular weight as mordant in electron microscopy. J. Cell Biol. 70:608-633. 20. Sprunt, K., W. Redman, and G. Leidy. 1968. Penicillin resistant alpha streptococci in pharynx of patient given oral penicillin. Pediatrics 42:957-975. 21. Thornsberry, C., T. L. Gavan, and E. H. Gerlach. 1978. Cumitech 6, New developments in antimicrobial susceptibility testing. Coordinating ed., S. Marcus and J. C. Sherris. American Society for Microbiology, Washington, D.C. 22. Tomasz, A. 1979. The mechanism of the irreversible antimicrobial effects of penicillins. Annu. Rev. Microbiol. 33:113-137. 23. Tomasz, A., J. D. Jamieson, and E. Ottolenghi. 1966. The

136 GUTMANN AND TOMASZ fine structure of Diplococcus pneumoniae. J. Cell Biol. 22:453-467. 24. Ubukata, K., M. Konno, and R. Fuji. 1975. Transduction of drug resistance to tetracycline, chloramphenicol, lincomycin and clindamycin with phage induced from Strepto- ANTIMICROB. AGENTS CHEMOTHER. coccus pyogenes. J. Antibiot. 28:681-688. 25. Zighelboim, S., and A. Tomasz. 1980. Penicillin-binding proteins of multiply antibiotic-resistant South African strains of Streptococcus pneumoniae. Antimicrob. Agents Chemother. 17:434-442.