Krzysztof Sieradzki and Alexander Tomasz* The Rockefeller University, 1230 York Avenue, New York, NY 10021, USA

Transcription

1 Journal of Antimicrobial Chemotherapy (1997) 39, Suppl. A, JAC Suppression of -lactam antibiotic resistance in a methicillin-resistant Staphylococcus aureus through synergic action of early cell wall inhibitors and some other antibiotics Krzysztof Sieradzki and Alexander Tomasz* The Rockefeller University, 1230 York Avenue, New York, NY 10021, USA We tested the effect of a number of mechanistically distinct antibacterial agents on the expression of methicillin resistance in a highly and homogeneously resistant strain of methicillin-resistant Staphylococcus aureus. The antibiotics, used at 0.25 x MIC, included inhibitors of early steps in peptidoglycan synthesis (fosfomycin, -chloro-d-alanine, D- cycloserine); bacitracin; teicoplanin and vancomycin; -lactam inhibitors chosen on the basis of their relatively selective affinities for penicillin-binding proteins 1, 2, 3 and 4 of S. aureus (imipenem, cefotaxime, cephradine and cefoxitin); compounds that inhibit various steps in protein synthesis (tetracycline, chloramphenicol, gentamicin, erythromycin and quinupristin/dalfopristin) and an inhibitor of DNA gyrase (temafloxacin). All inhibitors of early cell wall synthesis caused reduction of methicillin resistance and change from the homogeneous to the heterogeneous methicillin-resistant phenotype. Similar effects were obtained with only cephradine out of the four -lactams tested, and with erythromycin and quinupristin/dalfopristin as well. The other inhibitors of protein synthesis and DNA gyrase had no effect. Introduction Methicillin-resistant Staphylococcus aureus (MRSA) has become the highest-ranking pathogen worldwide. 1,2 In addition, a distinct shift in the direction of highly and homogeneously methicillin-resistant phenotypes has been observed among contemporary MRSA isolates, in contrast with historically earlier MRSA which often showed heterogeneous -lactam-resistant phenotypes. 3 Major MRSA clones defined by genetic techniques have also acquired resistance to numerous additional antibacterial agents. 2,4 It is conceivable that a combination of antibacterial agents with different modes of action would be useful against some of these strains. A combination of - lactamase inhibitors and penicillinase-sensitive penicillins was shown to be effective against several strains of MRSA which carried both the -lactamase as well as the meca-related antibiotic resistance mechanisms. 5 An additional reason for testing a combination of antibacterial agents is related to genetic studies in which Tn551 insertional mutants were shown to have reduced methicillin resistance. 6 In a number of these mutants the inactivated gene was involved with an early step in cell wall biosynthesis, such as extension of oligoglycine crosslinking peptides, amidation of D-glutamate -carboxyl groups, or the addition of the diamino acid residue to the stem peptide (for review see de Lencastre et al. 7 ). For these reasons we tested a panel of antibacterial agents targeted at various steps of bacterial metabolism for their effect on the expression of methicillin resistance on an MRSA strain of high and homogeneous methicillinresistant phenotype. In order to facilitate the interpretation of experiments we chose MRSA strain COL, which neither produces -lactamase nor has a functioning chromosomally located meci/mecr regulatory gene. 8,9 Materials and methods MRSA strain COL from the Rockefeller University collection was grown in tryptic soy broth (TSB) by aeration at 37 C and susceptibility to the various antibiotics and to their synergic combinations was determined by population analysis. 10 Freshly grown overnight cultures of bac- *Corresponding author. Tel: ; Fax: The British Society for Antimicrobial Chemotherapy 47

2 K. Sieradzki and A. Tomasz teria were plated at various dilutions on tryptic soy agar (TSA) containing two-fold dilutions of the antibacterial agents. Cell walls of the bacteria produced during growth in the presence of subinhibitory concentrations of various antibacterial agents were analysed by a high performance liquid chromatography (HPLC) method, 11 except that the step involving treatment with phosphatase was omitted. Antibacterial agents were obtained through the courtesy of pharmaceutical companies. Results All antibiotics tested except gentamicin and erythromycin gave homogeneous population analysis profiles (PAPs). On the basis of these PAPs the value of 0.25 MIC and occasionally 0.5 MIC of the particular antibacterial agent was chosen to test for possible effects on the expression of methicillin resistance in strain COL. Overnight cultures of the strain were placed at various dilutions on TSA agar containing a constant 0.25 MIC concentration of the antibiotic to be tested for synergy and two-fold dilutions of methicillin in the concentration range of 0.75 to 800 mg/l methicillin. Figure 1 is a composite of several such synergy experiments in which the agents tested at sub-mic concentration were inhibitors of early steps in cell wall biosynthesis. With each drug tested the homogeneous expression of methicillin resistance was changed to a unique heterogeneous phenotype, and the methicillin MIC of the majority cell population was reduced by various degrees. In the absence of the synergic agents the MIC of methicillin was 800 mg/l. This value was dramatically reduced to 6 mg/l (fosfomycin), 100 mg/l ( -chloro-d-alanine), 12 mg/l (D-cycloserine), mg/l (bacitracin), 50 mg/l (teicoplanin) and mg/l (vancomycin) in the presence of the particular synergic agents used. Figure 2 shows the effect of -lactam agents, inhibitors of various steps in protein synthesis and a DNA gyrase inhibitor on the expression of methicillin resistance. Only cephradine, erythromycin and quinupristin/dalfopristin caused reduction of methicillin resistance. The latter was the most effective in reducing the MIC of the majority of bacteria, from 800 to about mg/l. Imipenem, cefotaxime, cefoxitin, tetracycline, chloramphenicol, gentamicin and temafloxacin were ineffective under these test conditions. Figure 1. Effect of cell wall inhibitors on the phenotypic expression of methicillin resistance. Cultures of the MRSA strain COL grown in TSB were plated at different cell concentrations on TSA containing various concentrations of methicillin and 0.25 MIC of various inhibitors of cell wall synthesis. Colonies were counted after incubation at 37 C for 48 h and data are plotted to provide population analysis profiles (PAPS). (a) Fosfomycin, 15 mg/l; (b) -chloroalanine, 200 mg/l; (c) D-cycloserine, 15 mg/l; (d) bacitracin, 6 mg/l; (e) teicoplanin, 1.5 mg/l; and (f) vancomycin, 0.4 mg/l. PAP of strain COL in the absence of synergists ( ) and in the presence of synergists ( ). The asterisks indicate the agar plates from which colonies were picked to redetermine the MIC of the particular synergic agent (see Table). 48

3 Synergy and MRSA Figure 2. Effect of a variety of antibiotics on the phenotypic expression of methicillin resistance. Cultures of MRSA strain COL were grown and tested as described in the legend to Figure 1. Expression of methicillin resistance in the presence of sub-mic concentration (mg/l) of (a) imipenem (, 12), cefotaxime (, 200) and cefoxitin (, 100); (b) cephradine (, 50), (c) erythromycin (, 0.1); (d) quinupristin/dalfopristin (, 0.2), (e) gentamicin (, 0.1), tetracycline (, 50), chloramphenicol (, 1.5); (d) temafloxacin (, 0.05);, PAP of strain COL in the absence of the second drug. See the legend of Figure 1 for asterisk description. Colonies that retained their high degree of methicillin resistance even in the presence of synergically active antibacterial agents were picked from the plates containing 400 mg/l methicillin, 10 grown in liquid cultures, and Table. MIC values of antibiotics used as synergic agents. MICs were determined by population analysis for the parental strain COL and for the subpopulation of bacteria that retained high-level methicillin resistance even in the presence of the particular synergic agent ( homo population). Such colonies were picked from the population analysis plates containing 400 mg/l methicillin (see positions marked with an asterisk in Figures 1 and 2) MIC (mg/l) original homo Fosfomycin Chloro-D-alanine D-Cycloserine Bacitracin Teicoplanin Vancomycin Quinupristin/dalfopristin used to redetermine the MIC of the particular synergic agent. These colonies were referred to as homogeneous (homo) variants, following the original definition of this term. 12 The Table shows that the original MIC of the synergic agent was found to have increased in the homo colonies picked from the PAP plates. For instance, the MIC of fosfomycin, which was originally 60 mg/l against strain COL, was found to have increased to 500 mg/l in the colony picked from the area of high methicillin concentration in the PAP plate. Similarly, the original MIC of teicoplanin shifted from 3 6 up to 12 mg/l and that of vancomycin from 1.5 to 3 mg/l. Bacterial cultures grown in the presence of sub- MIC concentrations of the synergically active drugs were analysed to assess the effect of these growth conditions on the composition of cell wall peptidoglycan. Cell walls were prepared, peptidoglycan isolated and the family of muropeptides produced by enzymatic degradation was then separated by HPLC according to a recently described technique. 11 Figure 3 shows that the muropeptide composition of strain COL underwent major change when the bacteria were grown in the presence of 0.25 MIC of D-cycloserine. The appearance of a new peak (X) at a retention time of 1 2 min and the disappearance of muropeptide peak 1 are consistent with the 49

4 K. Sieradzki and A. Tomasz incorporation of the incomplete stem peptide (tripeptide) monomer into the cell wall of the bacteria exposed to D- cycloserine, a known inhibitor of the D-alanine ligase reaction. Bacteria grown in the presence of fosfomycin showed no detectable changes in the composition of peptidoglycan. Discussion Figure 3. Effect of growth in the presence of sub-mic of D- cycloserine on the muropeptide composition of staphylococcal cell walls. Cell wall peptidoglycan was isolated from strain COL grown in TSB (a) or in TSB supplemented with 15 mg/l D-cycloserine (b). Enzymatic peptidoglycan hydrolysates were prepared and analysed by the HPLC technique as described in Materials and methods. The observations described indicate that interference with early stages of peptidoglycan biosynthesis can drastically reduce methicillin resistance. In addition, bacteria treated with sub-mic concentrations of such inhibitors produced heterogeneous methicillin-resistance phenotypes characteristic of Tn551 insertional mutants of strain COL, and reminiscent of the phenotypes of historically early clinical isolates of MRSA. 3 The finding that every early inhibitor of cell wall synthesis was effective in producing these effects is consistent with the observations made with several of the Tn551 mutants, in which biochemical and genetic studies have identified the defective determinants as genes involved with early stages in cell wall biosynthesis. The finding that the colonies retaining the high degree of methicillin resistance in the presence of the synergic agent are actually mutants with increased MIC the synergist suggests that the so-called homo population of heterogeneous clinical strains, 12 and possibly the highly resistant subpopulations observed in Tn551 insertional mutants, may have a similar mechanism. We chose the -lactam inhibitors tested on the basis of their known relatively selective affinities for the penicillin-binding proteins of S. aureus. 13,14 The observation that imipenem, cefotaxime and cefoxitin had no detectable effect on the expression of methicillin resistance under the conditions tested suggests that penicillinbinding proteins (PBPs) 1, 2 and 4 (the selective targets of these three -lactams) are not involved in the expression of meca-dependent -lactam resistance. Interestingly, cephradine, a relatively selective inhibitor of PBP3, had an inhibitory effect at sub-mic concentration, raising the intriguing possibility that this particular PBP may be participating in some as yet undefined manner in the expression of methicillin resistance, perhaps through some cooperative functioning with PBP2A. While drugs that inhibit bacterial protein synthesis and DNA gyrase were without effect on the methicillinresistance phenotype, erythromycin and, to a larger degree quinupristin/dalfopristin, 15 caused considerable reduction in methicillin resistance as well as heterogeneous phenotype in our strain. We have no explanation for this finding, which may be related to the observation by Lorian et al. 16 of morphological changes in staphylococci treated with quinupristin/dalfopristin. We are not aware of any reports in the literature indicating involvement of a ribosomal step in bacterial cell wall biosynthesis. On 50

5 Synergy and MRSA the other hand, some of the recently identified auxiliary mutants were tentatively identified as genes encoding an alternative staphylococcal sigma factor, the inactivation of which blocked the expression of methicillin resistance. 1 7 It is conceivable that the complex multistep pathway connecting the meca gene and the methicillin-resistance phenotype also includes the expression of some protein co-factor essential for optimal expression of -lactam resistance through the meca-dependent resistance mechanism. References 1. Thornsberry, C. (1994). Epidemiology of staphylococcal infections a USA perspective. Journal of Chemotherapy 6, Suppl. 2, Vandenbroucke-Grauls, C. (1994). Epidemiology of staphylococcal infections a European perspective. Journal of Chemother - apy 6, Suppl. 2, Barber, M. (1964). Naturally occurring methicillin-resistant staphylococci. Journal of General Microbiology 35, Dominguez, M. A., de Lencastre, H., Liñares, J. & Tomasz, A. (1994). Spread and maintenance of a dominant methicillin-resistant Staphylococcus aureus (MRSA) clone during an outbreak of MRSA disease in a Spanish hospital. Journal of Clinical Micro - biology 32, Kobayashi, S., Arai, S., Hayashi, S. & Sakaguchi, T. (1989). In vitro effects of -lactams combined with -lactamase inhibitors against methicillin-resistant Staphylococcus aureus. Antimicrobial Agents and Chemotherapy 33, de Lencastre, H. & Tomasz, A. (1994). Reassessment of the number of auxiliary genes essential for expression of high-level methicillin resistance in Staphylococcus aureus. Antimicrobial Agents and Chemotherapy 38, de Lencastre, H., de Jonge, B. L. M., Matthews, P. R. & Tomasz, A. (1994). Molecular aspects of methicillin resistance in Staphylococcus aureus. Journal of Antimicrobial Chemotherapy 33, Murakami, K. & Tomasz, A. (1989). Involvement of multiple genetic determinants in high-level methicillin-resistance in Staph - ylococcus aureus. Journal of Bacteriology 171, Tesch, W., Ryffel, C., Strassle, A., Kayser, F. H. & Berger- Bächi, B. (1990). Evidence of a novel staphylococcal mecencoded element (mecr) controlling expression of penicillinbinding protein 2. Antimicrobial Agents and Chemotherapy 34, Tomasz, A., Nachman, S. & Leaf, H. (1991). Stable classes of phenotypic expression in methicillin resistant clinical isolates of staphylococci. Antimicrobial Agents and Chemotherapy 35, De Jonge, B. L. M., Chang, Y.-S., Gage, D. & Tomasz, A. (1992). Peptidoglycan composition of a highly methicillin-resistant Staphylococcus aureus strain: the role of penicillin binding protein 2A. Journal of Biological Chemistry 267, Hartman, B. J. & Tomasz, A. (1986). Expression of methicillin resistance in heterogeneous strains of Staphylococcus aureus. Antimicrobial Agents and Chemotherapy 29, Georgopapadakou, N. H., Smith, S. A. & Bonner, D. P. (1982). Penicillin-binding proteins in a Staphylococcus aureus strain resistant to specific -lactam antibiotics. Antimicrobial Agents and Chemotherapy 22, Chambers, H. F. & Sachdeva, M. (1990). Binding of -lactam antibiotics to penicillin-binding proteins in methicillin-resistant Staphylococcus aureus. Journal of Infectious Diseases 161, Low, D. E. (1995). Quinupristin/dalfopristin: spectrum of activity, pharmacokinetics, and initial clinical experience. Microbial Drug Resistance 1, Lorian, V., Amaral, L. & Fernandez, F. (1995). Post-antibiotic effect of RP59500 on Staphylococcus aureus defined by bacterial ultrastructure. Drugs in Experimental Clinical Research 11, Wu, S., de Lencastre, H. & Tomasz, A. (1996). Sigma-29S, a putative operon encoding an alternate sigma factor of Staphylo - coccus aureus: molecular cloning, DNA sequencing and role in stress response and antibiotic resistance. In Abstracts of the Ninety-Sixth General Meeting of the American Society for Micro - biology, New Orleans,