Two sets of Staphylococcus aureus isolates recovered from two patients exhibited similar susceptibility profiles except for oxacillin susceptibility (MSSA) or resistance (MRSA). SmaI macrorestriction and inter-IS256 PCR analysis showed patterns closely related to the Belgian epidemic MRSA clone 1 in each pair of MSSA/MRSA strains. Loss of one large SmaI DNA fragment and concurrent gain of a smaller fragment in the MSSA isolates was observed. The mecA sequence present in the MRSA was absent in the MSSA variant. Therefore, in vivo deletion of the mec region may occur in some lineages of S. aureus more frequently than previously thought.
Methicillin-resistant Staphylococcus aureus (MRSA) strains are common nosocomial pathogens that are also increasingly being encountered in community-acquired infections. Resistance to methicillin is determined by the mecA gene, which is present in a chromosomal insertion equal to or larger than 20 kb.1 While the precise origin of mecA remains unclear, several pathways for the evolution of methicillin-susceptible S. aureus (MSSA) strains into MRSA have been proposed.1 MSSA strains considered to be precursors of MRSA,2 as well as clinical isolates of MSSA and MRSA that are closely related genetically, have been described.3–5 However, the frequency of in vivo acquisition and/or deletion of the mec region by different S. aureus clones, and hence the proposed ‘clonality’ of MRSA,6 have remained an issue of debate. This study examined two sets of MRSA and MSSA strains, isolated consecutively or simultaneously from two patients.
Materials and methods
Antibiotics and susceptibility tests
Susceptibility to penicillin, erythromycin, clindamycin, spectinomycin, streptomycin, kanamycin, gentamicin, tobramycin, tetracycline, doxycycline, minocycline, ciprofloxacin, co-trimoxazole, rifampicin, fusidic acid, vancomycin and mupirocin was tested by a standard disc diffusion method with Neo-Sensitab tablets (Rosco, Taastrup, Denmark) using NCCLS breakpoints for susceptibility categorization. Oxacillin resistance was assessed by determining MICs using the Etest (AB BioDisk, Solna, Sweden) and a nuc/ mecA multiplex PCR as previously described.7 Resistance to cadmium chloride and mercuric chloride was tested as described.3
Molecular typing and mecA hybridization
Genomic macrorestriction (SmaI and SstII) followed by pulsed-field gel electrophoresis (PFGE) and inter-IS256 spacer polymerase chain reaction (PCR) were performed as described previously.8,9 Transfer of the SmaI DNA fragments and hybridization using a 32P-labelled 533 bp mecA probe (X52593: nucleotide positions 1416–1950) were carried out as described previously.4
Results and discussion
Two sets of MRSA and MSSA isolates, each obtained from a different patient, were studied. Isolates 91 (MRSA) and 92 (MSSA), were recovered consecutively from superficial wounds and decubitus ulcers over a 6 month period (in January and July 1996, respectively) in a 64 year old patient with Alzheimer's disease (patient 1) hospitalized in a chronic care unit, in André Vésale Hospital (Montigny-Le-Tilleul, Belgium). With the exception of oxacillin, their antibioticsusceptibility profiles were identical (resistance to erythromycin, clindamycin, spectinomycin, streptomycin, kanamycin, gentamicin, tobramycin, tetracycline and ciprofloxacin). Between isolation of the two strains, the patient had received several courses of piperacillin–tazobactam, pefloxacin and vancomycin for presumed sepsis.
Isolates 53 (MRSA) and 57 (MSSA) were isolated simultaneously, in July 1997, from a sputum sample from a 28 year old patient (patient 2) with cystic fibrosis, treated as an outpatient at Erasme Hospital (Brussels, Belgium). With the exception of oxacillin, their antibiotic susceptibility profiles were identical (resistance to erythromycin, clindamycin, spectinomycin, streptomycin, kanamycin, gentamicin, tobramycin, tetracycline, fusidic acid and ciprofloxacin). The patient had been chronically colonized with MRSA in his sputum since 1993, for which he had received several courses of oral fusidic acid, oral minocycline and tobramycin aerosol therapy. On three occasions in 1994 and 1995, MSSA had been reported in his sputum, with a resistance phenotype to other antimicrobials identical to that of the MRSA isolates. Between 1997 and 1999, only the MRSA variant was isolated from monthly sputum cultures.
By inter-IS256 PCR fingerprinting (not shown) and PFGE, each MSSA isolate was identical to its corresponding MRSA, except for the loss of a SmaI DNA macrorestriction fragment of either 208 or 220 kb and the concurrent gain of a smaller fragment (175 kb) [Figure (a)]. The two sets of strains were clonally related to each other and to the Belgian epidemic MRSA clone 1,8 their SmaI profiles differing by three to four DNA fragments [Figure (a)]. The hypothesis that a fragment containing the mecA gene was deleted in each case, was tested by hybridization with a probe spanning nucleotides 1416–1950 of the mecA gene. Strain NCTC 8325 was used as a negative control (Figure, lanes M), and representative strains of the two predominant Belgian epidemic MRSA clones 1 and 2,8 as positive controls (Figure, lanes 1 and 2, respectively). In both MRSA strains 53 and 91, the mecA sequence was present in the DNA fragment that was absent from the corresponding MSSA. This suggested that MRSA isolates 53 and 91 (Figure, lanes 3 and 5) had undergone mec region deletions of c.33 and 45 kb to yield MSSA isolates 57 and 92 (Figure, lanes 4 and 6), respectively. In contrast to the co-deletion of genes encoding resistance to methicillin, tetracycline and heavy metals reported by Inglis,3 resistance to antibiotics and heavy metals was identical in the MRSA/MSSA pairs studied here, suggesting that the resistance genes flanking the mecA gene were not deleted, a finding similar to that reported by Lawrence et al.4 However, this did not appear to correlate with the estimated size of the chromosome deletion based on PFGE. Possible explanations for this discrepancy include: (i) masking of the deletion phenotype by the presence of multi-copy resistance genes in other chromosomal regions/plasmids and (ii) overestimation of deletion size due to non-detection by PFGE analysis of small (<10 kb) restriction fragments of MSSA variants.
The observation that MSSA strains 92 and 57 were not recovered on subsequent microbiological sampling could be explained either by a decreased colonization ability of the MSSA variants or by the failure of isolation of a persistent but smaller MSSA population.
Instability of the S. aureus mec determinant has been reported in a limited number of clinical settings to date.2–5 Homologous MRSA and MSSA isolated from different patients in the same hospital or region have been described,3,5 while replacement of MRSA by MSSA was described after discontinuation of antibiotic therapy in a newborn.4 In our patients, intermittent or simultaneous isolation of both MRSA and MSSA variants led to uncertainty about treatment and infection control requirements. MRSA isolates recovered subsequently from patients 1 and 2 showed macrorestriction patterns identical to those of MRSA 91 and MRSA 53, respectively. MRSA variants persisted either in the absence of antibiotic therapy (patient 1) or during intermittent therapy with drugs (patient 2) that were equally active against MRSA and MSSA variants. Therefore, selection by antimicrobial therapy does not explain the apparent eradication of the MSSA variants.
The clonal relatedness of the isolates reported here to the widespread Belgian epidemic MRSA clone 1 supports the hypothesis that deletion of the mec element occurred in these patients, rather than insertion into MSSA from a putative donor like Staphylococcus epidermidis. Such a duplicate observation also suggests that this particular MRSA lineage possesses a less stable mec insertion than other MRSA clones. The frequency of mec deletion in vivo may be underestimated by the common practice of performing susceptibility testing on a sweep of colonies, thereby possibly masking the co-existence of mixed susceptibility phenotypes. The isolation of genomically related MSSA and MRSA in the two patients reported here suggests that in vivo acquisition and/or deletion of the mec region may not be as rare a phenomenon as previously believed.
This study was initiated during a working visit by P.T.T., supported by a FEMS Fellowship for Young Scientists, to the Department of Microbiology, Erasme Hospital.