Introduction

Methicillin-resistant Staphylococcus aureus (MRSA) infections have increased dramatically during the last several decades and have been shown to be associated with higher mortality and costs than those caused by methicillin-sensitive S. aureus (MSSA) isolates [1]. In the last decade, community-acquired MRSA (CA-MRSA) has emerged worldwide [2]. CA-MRSA can be distinguished from hospital-acquired MRSA (HA-MRSA) by its susceptibility to many non-β-lactam antibiotics, the presence of staphylococcal cassette chromosome mec (SCCmec) types IV and V (as opposed to types I–III), and lack of previous patient exposure to a healthcare setting. Many CA-MRSA strains have been reported to carry the Panton–Valentine leukocidin (pvl) genes and to cause invasive infections associated with increased mortality. Yet this last characteristic is not required to define CA-MRSA [2–4].

The objectives of this study were to determine the prevalence of CA-MRSA carriage among young children and their parents in Israel and to determine the genetic relatedness of these isolates to other S. aureus strains isolated from this population.

Patients and methods

The study was approved by the Sheba Medical Center Ethics Committee. Informed consent was obtained from all patients or their parents or guardians. During February 2002 and October 2002, children aged ≤40 months and their accompanying parents or guardians, who were visiting any of 53 participating primary-care pediatric clinics, were enrolled. The clinics were located in central Israel and belonged to a major health management organization (Maccabi Healthcare Services). Each participant was included only once.

Nasal swabs were obtained from each participant and data regarding demographic characteristics, possible contact with a healthcare facility, the physician’s diagnosis of the child on the specific visit, and prior antibiotic treatment, were collected using questionnaires and patient files. S. aureus was identified by morphology, β-hemolysis, catalase, DNAase, and coagulase production. Antibiotic susceptibilities were determined using the VITEK-2 system with plate AST P536 (bioMérieux, Hazelwood, MO, USA). Oxacillin resistance was verified with MRSA screening agar, the broth microdilution MIC technique [5], E-test (AB Biodisk, Solona, Sweden) and PCR for the presence of the mecA gene (Promega, Madison, WI, USA). S. aureus isolates identified in children and their parents (150 isolates) were further analyzed by pulsed-field gel electrophoresis (PFGE) (New England BioLabs, Boston, USA) [6]. SCCmec typing [7] and multilocus sequence typing (MLST) [8] were assessed for all MRSA isolates and for the common MSSA clone. The following virulence factors were detected by PCR [9]: the Panton-Valentine leukocidin (pvl) locus (lukS-PV-lukF-PV), sea, seb, sec, sed, see, seh, sej, sel, sep, egc, tst, eta, etb, etd, lukE-lukD, lukM, hlb, hlg, and hlg2, edin, and agr genes. Doubling-times for the five MRSA isolates and the predominant MSSA clone were compared [6].

Results and discussion

A total of 3,373 persons were sampled, including 1,768 children (55% male, median age 1.4 y, range 14 days–40 months) and 1,605 parents (86.7% mothers, median age 30 y, range 18–55 years). S. aureus was isolated from 201 (11.4%) children and from 378 (23.6%) parents. A predominant MSSA clone, exhibiting a common pulsotype, designated clone F, was found in 25% of 150 isolates tested by PFGE. Five MRSA isolates carrying the mecA gene were detected, two (1%) in children and three (0.8%) in parents. The MRSA carriage rate was thus 1.1/1,000 in children and 1.8/1,000 in their parents. Three MRSA isolates had a multidrug-resistant (MDR) phenotype (resistance to ≥3 antimicrobial groups). The carriers of these three isolates reported recent contact with a healthcare facility. The carriers of the two non-MDR MRSA isolates were unrelated adults (mothers of non-carrier children), who lived in different cities and who had no apparent contact with each other or with any healthcare facility (Table 1). All five MRSA carriers were healthy on the day of MRSA detection and they had not received antibiotic treatment within the prior month. The two non-MDR MRSA isolates possessed SCCmec type IV and were classified as MLST type 45 (ST45-MRSA-IV). Their PFGE pattern was similar to that of the common MSSA isolate, clone F, but their genotype differed in two bands; thus, they were subtyped as F1 and F2 (Fig. 1). The virulence profile was mostly similar to that of the common MSSA, but ST45-MRSA-IV carried only one enterotoxin gene, as compared to four, carried by the ST-45-MSSA strain. The three MDR-MRSA isolates had distinct PFGE patterns, all of which differed from the community S. aureus patterns observed and were similar to HA-MRSA clones (Fig. 1, Table 1).

Table 1 Characteristics of the MRSA strains and the common MSSA strain isolated in this study
Full size table
Fig. 1
figure 1

Pulsed-field gel electrophoresis (PFGE) pattern following SmaI restriction of the MRSA and MSSA clones isolated from the community. Lane 1 lambda DNA ladder molecular-weight marker; lanes 2–4 HA-MRSA clones (clones X, Y, Z); lanes 5 and 6 CA-MRSA ST45-MRSA-IV strains (subtypes F1 and F2) possessing a subtype of the common ST 45-MSSA clone — F (lanes 7–15)

Full size image

The pvl locus was not carried by any of the CA-MRSA isolates. These CA-MRSA isolates grew significantly faster than the genetically related community MSSA isolates (Table 1) and at a similar rate to two of the MDR-MRSA isolates.

The prevalence of CA-MRSA is increasing, but it varies among different countries [2]. Several reports have related the rapid spread to a rapid growth rate [10], and to super-adapted strains, due to the combination of the pvl determinant with the mecA gene [2]. In our study population we found a low prevalence of MRSA carriage (15/10,000 population), and an even lower “true” CA-MRSA rate.

There is controversy regarding the origin of CA-MRSA strains [11–13]. Recent evidence supports a rapid evolution of CA-MRSA from common MSSA [14, 15]. The two CA-MRSA isolates described in this study were genetically related to the common MSSA clone (ST45-MSSA) identified in this population, thus supporting the idea of strain evolution. These isolates could either represent two independent incidents of mecA gene acquisition or, conversely, spread of a single clone, which could represent a hospital escape. Both scenarios are equally worrisome.

The strain type ST45-MRSA-IV, described herein, differs from the commonly reported CA-MRSA in that it does not carry the pvl gene or agr type 3. ST45 was previously found in both MRSA and MSSA strains [11, 15]. ST45-MRSA, also known as the Berlin epidemic MRSA or USA600 strain, has been shown to carry different SCCmec types (II and IV). Recently, Wannet et al. [15] reported that ST45-MRSA has rapidly emerged in the Netherlands, where it was first detected in 2000 and became endemic by 2002. They found four different SCCmec types carried by this clone, and six of the SCCmecIV isolates were from different geographical regions, thus suggesting that multiple introductions of SCCmec into MSSA have occurred. We recently reported an outbreak caused by this particular strain (non-MDR, SCCmec IV, pvl-negative, ST-45-MRSA), in a neonatal ICU [4]. This implies that although it is rare in the community (and previously not detected in this hospital), this strain has the potential to spread rapidly and to cause lethal infections. Furthermore, the rapid in vitro growth of these isolates compared to their suspected progenitor MSSA strain, suggests a competitive advantage for rapid spread. Interestingly, the CA-MRSA strain also differed from the ST-45-MSSA strain in its enterotoxin gene profile. The significance of this observation has yet to be determined.

The results of this study should be interpreted with respect to two components of its design: (1) the population was sampled from primary-care clinics, which may not accurately represent the healthy population; (2) the reference MSSA strain was obtained from children and their parents, who both carried S. aureus. This strain might not be fully representative of the common MSSA strain(s) in the general population. Further surveillance and continuous follow-up is required in order to determine the actual rate at which such strains are evolving as well as their pathogenic potential.