Evolution and diversity of community-associated methicillin-resistant Staphylococcus aureus in a geographical region

Based on phenotypic and genotypic typing methods, community onset methicillin-resistant Staphylococcus aureus infections are caused by healthcare-associated MRSA (HA-MRSA) strains, which appear to have been transferred from hospitals or healthcare facilities into the community by patients or healthcare workers [1], or by community-associated MRSA (CA-MRSA) strains, which have been isolated from people who have had little or no contact with healthcare facilities or healthcare workers [2]. This distinction between community and healthcare facility however has become blurred with the replacement of HA-MRSA with CA-MRSA in hospitals [3, 4].

In contrast to HA-MRSA, CA-MRSA strains are generally more susceptible to non beta-lactam antibiotics, grow significantly faster, have different clonal backgrounds, carry smaller staphylococcal cassette chromosome mec (SCCmec) elements (most commonly SCCmec type IV or type V), have enhanced virulence properties and frequently harbor genes expressing Panton-Valentine leukocidin (PVL) [5–8]. Rather than a worldwide spread of a single clone multiple CA-MRSA clones have emerged from diverse genetic backgrounds. Several well characterized CA-MRSA clones predominate in different regions: Sequence type (ST) 8-IV [2B] (USA300) and ST1-IV [2B] (USA400) in North America [9, 10]; ST80-IV [2B] (European clone) in Europe [8], North Africa [11] and the Middle East [12]; ST59-V [5C2&5] (Taiwan clone) in Taiwan [13]; ST93-IV [2B] (Queensland clone) in Australia [14], ST30-IV [2B] (South West Pacific [SWP] CA-MRSA) in the Western Pacific [15, 16], and ST772-V [5C2] (Bengal Bay clone) in India and Bangladesh [17]. Transmission of these clones into other regions has occurred [18, 19]. This occurrence of concurrent epidemics of CA-MRSA in many countries by different clones has been striking. Equally noteworthy are a number of common features of these epidemics, prominent among them the ability to cause severe infections in young otherwise healthy people and the carriage of the lukF-PV/lukS-PV PVL encoding genes by the organism.

The earliest report of CA-MRSA infections involved indigenous people living in remote communities in the sparsely populated Kimberley region of Western Australia (WA) [20]. Approximately 50% of the people in this region are indigenous, many of whom live in poor socioeconomic conditions. Infected skin lesions and staphylococcal sepsis occur frequently and empirical antistaphylococcal therapy is often prescribed. Colloquially known as "WA-MRSA", the early isolates have a similar pulsed-field gel electrophoresis (PFGE) pattern and have subsequently been characterized as a single clone; PVL-negative WA5 (ST8-IV/spa t008) [21]. By 2006 22 CA-MRSA clones were identified in WA, with PVL-negative WA 1 (ST1-IV [2B]/t127) replacing WA5 as the predominant clone [22]. At this time CA-MRSA from indigenous people living in remote areas outside of WA were reported in the Northern Territory [23], Queensland [24] and Central Australia [25]. As may be expected in a geographically large country with relatively few dense concentrations of population, often separated by large areas of desert, different CA-MRSA clones evolved in these communities.

In 1982 colonization or infection with MRSA became a notifiable condition in WA. For infection control purposes all MRSA isolated in the state since 1997 have been referred to the Australian Collaborating Centre for Enterococcus and Staphylococcus Species (ACCESS) Typing and Research where based on molecular markers they are characterized as either HA-MRSA or CA-MRSA [26]. Although a state-wide policy of screening all patients and healthcare workers who have lived outside the state for MRSA has prevented HA-MRSA from becoming endemic in Western Australian hospitals, it has not prevented CA-MRSA from becoming established in the community. In WA the public health system is divided into two metropolitan health regions and seven country health regions. The state encompasses an area of 1.02 million square miles and has a population of approximately 2.24 million people. In 1983, the overall rate of MRSA notifications was 10 per 100,000 persons in the rural country health regions and 7/100,000 in the metropolitan regions [27]. By 2006 notifications rates throughout the state had increased to 179/100,000 persons of which 144/100,000 were CA-MRSA. In the metropolitan health regions the CA-MRSA notification rate was 134/100,000 whilst in the Kimberley health region the CA-MRSA notification rate had increased 40-fold to 391/100,000 [18].

CA-MRSA is thought to emerge when a locally prevalent strain of methicillin susceptible S. aureus (MSSA) acquires a SCCmec element and utilizes mobile genetic elements and single nucleotide polymorphisms to establish local and geographic niches [28]. As WA is a remote region in which all MRSA isolates are referred to a central typing laboratory it is an ideal environment to study the emergence and evolution of CA-MRSA. MLST, SCCmec, spa typing and microarray DNA is performed on all isolates with a unique PFGE pulsotype. The aim of this study is to determine the genetic relatedness of WA CA-MRSA clones within different MLST clonal clusters (CC) providing an insight into the frequency of S. aureus SCCmec acquisition within a region. The genetic profile of these clones may also offer an explanation why only a few WA CA-MRSA clones have successfully adapted to the community environment.

The 83 unique PFGE strains isolated in Western Australia from 1989 to 2010 were nuc and mecA gene positive by PCR. The DNA microarray S. aureus species markers gapA (glyceraldehyde 3-phosphate dehydrogenase) and rrn STAU (S. aureus ribosomal marker) were detected in all strains. The array's linear primer elongation method detected the katA (catalase A), coA (coagulase), nuc, spa (protein A) and sbi (IgG-binding protein) S. aureus species markers in 78 strains. These markers were either not detected or detected only by random amplification in five strains (WA8, WA47, WA72, WA76 and WA79).

Forty six STs were identified by MLST. Using the MLST website's eBURST V3 algorithm 45 STs were grouped into 18 CCs and two singletons (Figure 1). The CC for WA76 (ST1303) has not been determined.

Several SCCmec types and subtypes, novel SCCmecs, and composite SCCmecs were identified. Forty five strains harbor SCCmec IVa-d [2B] (31 IVa, 2 IVb, 9 IVc, 3 IVd), 12 strains SCCmec V [5C2] and two strains SCCmec VIII [4A]. Two strains have non typeable SCCmec IV subtypes and four strains have a SCCmec element with a novel ccr gene complex including three with a class B mec gene complex and one with a class A mec complex. Eighteen strains harbor SCCmec elements with composite ccr gene complexes including 12 with SCCmec V [5C2&5] (5C2 plus ccrC1 allele 8), three with SCCmec IVa [2B]&5 (2B plus a type 5 ccr gene complex), one with V (5C2)&2 (5C2 plus a type 2 ccr gene complex) and two with V [5C2&5]&2 (a composite SCCmec V element plus a type 2 ccr gene complex).

The MLST, spa type, agr type, capsule type, SCCmec, antibiogram, resistance genotype, lukF/S-PVL genes, enterotoxin genes and bacteriophage associated virulence genes of each unique PFGE strain are provided in Additional File 1. Information on target genes, probes, and primers is provided in Additional File 2. Complete hybridization profiles for the individual strains can be provided on request.

As all MRSA isolated in WA are referred to a central typing laboratory it is possible to investigate the emergence and evolution of CA-MRSA in a remote region.

Prior to the global evolution and expansion of CA-MRSA, five CA-MRSA clones were identified in the indigenous population living in the remote communities of the sparsely populated Kimberley, Pilbara and Eastern Goldfield regions of WA [29]. These five PVL negative clones include: WA1 (CC1: ST1-IVa [2B]/t127), WA2 (CC88: ST78-IVa [2B]/t3205), WA3 (CC5: ST5-IVa [2B]/t002), WA4 (CC45 ST45-V (5C2)/t123) and WA5 (CC8: ST8-IVa [2B]/t008). WA5 and WA1 were originally isolated from clinical specimens in 1989 and 1995 respectively, and WA2, WA3 and WA4 from nasal carriage specimens in 1995. The emergence of CA-MRSA clones in different MLST clonal clusters indicates horizontal transmission of the SCCmec element into S. aureus has occurred on at least five occasions in these remote communities: SCCmec IVa [2B] into CC1 (ST1), CC5 (ST5), CC8 (ST8), CC88 (ST78), and SCCmec V [5C2] into CC45 (ST45). Based upon the spa type and the DNA microarray profile at least six evolutionary events have occurred on at least three occasions from these clones (ie vertical transmission of the SCCmec element): twice from WA1, WA3 and WA5 (Figure 2). Vertical transmission of the SCCmec element has not been identified for WA4 or WA2.

The emergence of WA1, WA2 and WA3 has been due to the acquisition and insertion of the small and highly mobile type IVa [2B] SCCmec element, presumably harbored by methicillin resistant coagulase negative staphylococci (MRCNS). Several hypotheses to explain the transmission of a SCCmec element from MRCNS to S. aureus have been proposed including the increased use of antimicrobials within a community [35]. Many of the Kimberley indigenous population live in poor socioeconomic conditions. Staphylococcal skin lesions, commonly resulting from scabies infestation, trachoma and venereal diseases such as chlamydia and gonorrhea occur frequently in this population. Consequently empirical therapy using β-lactamase stable penicillins and azithromycin is often prescribed [36]. The frequent use of these antimicrobials may have assisted in the acquisition of the SCCmec element and erm genes into S. aureus. Genetic studies however have shown these newly emerged CA-MRSA clones did not originate in the predominant methicillin-susceptible S. aureus (MSSA) clones found in these communities, suggesting not all clones are able to acquire or retain the SCCmec element [37]. The subsequent dissemination of WA1, WA2 and WA3 into the wider community suggests the acquisition of the SCCmec element and the erm genes has given these clones a selective advantage. WA4 and WA5 however have not been successful in spreading beyond the indigenous communities suggesting the acquisition of the SCCmec element does not provide a universal selective advantage.

Many of the remaining 46 CA-MRSA clones, identified between July 2003 and June 2010, were not isolated in remote WA indigenous communities. The geographical spread of CA-MRSA over long distances and across cultural borders is believed to be a rare event compared to the frequency in which the SCCmec element is acquired by S. aureus [38]. Most of these clones are therefore likely to have evolved in WA. Some clones are slvs and dlvs of pre-existing CA-MRSA, and their SCCmec type, spa type and DNA microarray profile suggests vertical transmission of the SCCmec element has occurred. However the emergence of MRSA in several unrelated S. aureus clonal clusters suggests horizontal transmission of the SCCmec element has also occurred. SCCmec typing and spa typing and DNA microarray results also suggests horizontal transfer of SCCmec elements has occurred into the same CC on more than one occasion.

Although several SCCmec elements have been acquired by multiple S. aureus clones from which many CA-MRSA clones have emerged, only a few clones have successfully adapted to the WA community environment. Between July 2009 to June 2010 4,691 MRSA were referred to ACCESS Typing and Research of which 3,931 were characterized as CA-MRSA. Overall 84% (3,024) of isolates were from clinical infections and the 16% (907) from colonized patients. Approximately 88% of CA-MRSA were identified as WA1 (40%), WA2 (24%) and WA3 (8%). For most clones, including WA4 and WA5 only a few isolates were detected. (http://www.public.health.wa.gov.au/3/896/3/camrsa.pm).

For many slv and dlv CA-MRSA only a small number of isolates have been detected suggesting changes in the housekeeping genes may have conferred a fitness cost or did not allow the SCCmec element to be maintained. For example WA45 and WA57 are slvs of ST1 and their SCCmec and spa type and DNA microarray profile suggest they have evolved from WA1 (Figure 2). WA45 was first identified in 2006 and WA57 in 2007. Although WA1 has become the most successful CA-MRSA clone in the WA community only one isolate of WA45 and two isolates of WA56 have so far been identified (http://www.public.health.wa.gov.au/3/896/3/camrsa.pm).

Six PVL positive pandemic CA-MRSA clones (plus three closely related clones) have been isolated in WA: Bengal Bay CA-MRSA (ST772-V [5C2]/t3387), USA300 MRSA (ST8-IVc [2B]/t008), SWP CA-MRSA (ST30-IVc [2B]/t019), Taiwan CA-MRSA (ST59-V [5C2&5]/t437 and the slv ST952-V [5C2&5]/t1950), European CA-MRSA (ST80-IVc [2B]/t044 and the slvs, ST583-IVc [2B]/t044 and ST728-IVc [2B]/t044), and the Queensland CA-MRSA (ST93-IVa [2B]/t202). The epidemiology of the USA300 and Taiwan CA-MRSA clones in WA and the Queensland and SWP CA-MRSA clones in Australia have previously been reported [18, 31, 32]. Patients colonized or infected with the Bengal Bay clone have been observed to be epidemiologically linked to Indian healthcare workers (unpublished data). The USA300, European, Taiwanese and Bengal Bay CA-MRSA clones are not frequently isolated in WA. This may be due, in part, to WA Health Department infection control interventions applied to patients who are colonized or infected with international PVL positive pandemic clones. A seventh pandemic clone has recently been identified. The DNA microarray profile and the SCCmec element of the PVL negative ST398-V [5C2&5] is indistinguishable from the pandemic ST398 clone initially isolated from pigs and pig farmers in the Netherlands [39]. Only one isolate, from a patient with travel outside of Australia, has been identified in WA.

The Queensland clone (ST93-IVa [2B]) first detected on the east coast of Australia in the Caucasian population in 2000 [40], has become one of the most prevalent CA-MRSA isolated in Australia [18] and in 2010 accounted for 18% of CA-MRSA in WA. This suggests the acquisition of the SCCmec element has given this clone a selective advantage. Although the Queensland clone is believed to have been introduced into WA in 2001 [22], PVL positive ST93-MSSA was identified as the most prevalent S. aureus clone in WA's remote indigenous communities in surveys performed in the mid 1990s. Although found in an environment of high β-lactam use a methicillin-resistant variant of ST93-MSSA was not found in WA during these surveys.

WA1, WA2 and WA3 are PVL negative and do not harbor multiple virulence genes (Tables 1). Similarly the successful Queensland clone, although PVL positive, carries almost no other exotoxin genes and no additional resistance genes. Although most other WA CA-MRSA clones are also PVL negative, many of these clones have acquired multiple resistance and/or virulence determinants (Tables 1). For example WA78 (ST188-IVa [2B]/t315) in addition to mecA and blaZ, harbors aacA-aphD, tetK and cat and is phenotypically resistant to erythromycin, trimethoprim and ciprofloxacin; WA64 (ST5-IVa [2B]/t3778) has acquired seA enterotoxin genes and edinA and lukF-PV lukS-PV virulence genes; and WA62 (ST923[ST8slv]-IVa [2B]/t1635) harbors seD+seJ+seR and seK+seQ enterotoxin genes and lukF-PV lukS-PV. The acquisition of multiple resistance and/or virulence genes may have come at a high fitness cost as none of these clones have established a niche in the WA community.

As WA1, WA2 and WA3 CA-MRSA lack PVL as well as other virulence genes that are found in pandemic international CA-MRSA clones, such ACME in USA300, the epidemiology of CA-MRSA disease in WA is different to other regions. Outside of WA the majority of diseases related to CA-MRSA infection are severe skin and soft tissue infections such as soft tissue abscess, carbuncles and furuncles. Many of these infections have occurred in healthy individuals, especially children and adolescents, usually via skin-to-skin contact [41]. In WA the majority of CA-MRSA related diseases were initially associated with the indigenous population and then other groups normally susceptible to S. aureus infections such as the elderly. As the original WA CA-MRSA are PVL negative, many of these infections were superficial skin infections such as impetigo. However with the introduction of the PVL-positive Queensland CA-MRSA clone more severe skin and soft tissues infections have been observed.

The limitation of this study is that only the initial isolate of each PFGE pulsotype was included in the study. To determine if the successful CA-MRSA clones found in the WA community are evolving the genetic profiles of subsequent isolates need to be investigated.

In conclusion although the vertical and horizontal transmission of SCCmec elements into S. aureus has occurred on multiple occasions in the WA community only three WA CA-MRSA clones have found an ecological niche. These three PVL negative clones harbor few additional resistance and virulence genes which paradoxically may account for their success.