The molecular epidemiology of the highly virulent ST93 Australian community Staphylococcus aureus strain.

In Australia the PVL-positive ST93-IV [2B], colloquially known as "Queensland CA-MRSA" has become the dominant CA-MRSA clone. First described in the early 2000s, ST93-IV [2B] is associated with skin and severe invasive infections including necrotizing pneumonia. A singleton by multilocus sequence typing (MLST) eBURST analysis ST93 is distinct from other S. aureus clones. To determine if the increased prevalence of ST93-IV [2B] is due to the widespread transmission of a single strain of ST93-IV [2B] the genetic relatedness of 58 S. aureus ST93 isolated throughout Australia over an extended period were studied in detail using a variety of molecular methods including pulsed-field gel electrophoresis, spa typing, MLST, microarray DNA, SCCmec typing and dru typing. Identification of the phage harbouring the lukS-PV/lukF-PV Panton Valentine leucocidin genes, detection of allelic variations in lukS-PV/lukF-PV, and quantification of LukF-PV expression was also performed. Although ST93-IV [2B] is known to have an apparent enhanced clinical virulence, the isolates harboured few known virulence determinants. All PVL-positive isolates carried the PVL-encoding phage ΦSa2USA and the lukS-PV/lukF-PV genes had the same R variant SNP profile. The isolates produced similar expression levels of LukF-PV. Although multiple rearrangements of the spa sequence have occurred, the core genome in ST93 is very stable. The emergence of ST93-MRSA is due to independent acquisitions of different dru-defined type IV and type V SCCmec elements in several spa-defined ST93-MSSA backgrounds. Rearrangement of the spa sequence in ST93-MRSA has subsequently occurred in some of these strains. Although multiple ST93-MRSA strains were characterised, little genetic diversity was identified for most isolates, with PVL-positive ST93-IVa [2B]-t202-dt10 predominant across Australia. Whether ST93-IVa [2B] t202-dt10 arose from one PVL-positive ST93-MSSA-t202, or by independent acquisitions of SCCmec-IVa [2B]-dt10 into multiple PVL-positive ST93-MSSA-t202 strains is not known.

Introduction

The community-associated methicillin resistant Staphylococcus aureus (CA-MRSA) worldwide epidemic is polyclonal, however several well characterized clones predominate in different regions of the world: Sequence type (ST) 8-IV [2B] (USA300) and ST1-IV [2B] (USA400) in North America [1], [2]; ST80-IV [2B] (European clone) in Europe [3], North Africa [4] and the Middle East [5]; ST59-V [5C2&5] (Taiwan clone) in Taiwan [6], ST30-IV [2B] (South West Pacific [SWP] CA-MRSA) in the Western Pacific [7], [8] and ST772-MRSA-V [5C2] (Bengal Bay clone) in India and Bangladesh [9]. Transmission of these clones into other regions has occurred [10], [11]. The 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 lukS-PV/lukF-PV Panton Valentine Leukocidin (PVL) encoding genes by the organism.

In Australia the PVL - positive ST93-IV [2B], colloquially known as “Queensland CA-MRSA”, has recently emerged to become the dominant CA-MRSA clone. First described in the early 2000s, ST93 is a singleton by MLST eBURST analysis and is therefore distinct from other S aureus clones [12].

In the 2010 Australian Group for Antimicrobial Resistance (AGAR) Community S aureus Surveillance Programme (SAP10) ST93-IV [2B] accounted for 41.4% of all CA-MRSA, 27.6% of all MRSA and 4.9% of all S aureus community-onset infections (http://www.agargroup.org/files/FED%20REPORT%20SAP210%20MRSA%20FINAL%20shrink.pdf.). The mean age of patients with ST93-IV [2B] infections (31 years, median 25 years) was significantly lower (P<0.0001) than the mean age of patients with PVL negative CA-MRSA infections (53 years; median 57 years).

ST93-IV [2B] is associated with skin infection and severe invasive infection including necrotizing pneumonia, deep-seated abscess, osteomyelitis, septic arthritis and septicaemia [13], [14], [15]. Although ST93-IV [2B] has an apparent enhanced clinical virulence, the recently sequenced ST93-IV [2B] strain “JKD6159” has a relative paucity of recognizable virulence determinants [16], [17]. This strain however does contain genes encoding three important CA-MRSA virulence factors, Hla, PVL and α-type phenol soluble modulins (PSMs), and when compared to three other well-characterised Australian MRSA strains, ST1-IV [2B], ST30-IV [2B] and ST239-III [3B] and the epidemic North American strain, USA300, was shown to be the most virulent in two in vivo models [17].

While predominately an Australian strain, ST93-IV [2B] has been reported in New Zealand, accounting for 5.1% of all MRSA referred to the Institute of Environmental Science and Research in 2010 (http://www.surv.esr.cri.nz/PDF_surveillance/Antimicrobial/MRSA/aMRSA_2010.pdf), and in the United Kingdom, [18], where many cases have epidemiological links to Australia.

In Western Australia (WA) ST93-IV [2B] was first identified in 2003 [19] and in SAP10 accounted for 28.8% of the state’s CA-MRSA community-onset infections. In the mid 1990s S aureus screening of indigenous people living in WA remote communities demonstrated the most prevalent methicillin susceptible S aureus (MSSA) linage isolated was the PVL-positive ST93 MSSA clone [20]. Although seven CA-MRSA clones from genetically diverse backgrounds were identified in these communities, no ST93 MRSA was found during this time.

As Australia is a geographically large country with the majority of the population densely concentrated in a few major cities which are separated in many instances by vast desert areas, it is to be expected that different CA-MRSA clones will have evolved in different areas of Australia. To better understand the molecular epidemiology of ST93-IV [2B], the aim of this study was to analyse the genetic relatedness of S. aureus ST93 isolated throughout Australia over an extended period and to determine if the increased prevalence of ST93-IV [2B] has been due to the widespread transmission of a single strain of ST93-IV [2B] or has been due to multiple independent acquisitions of the SCCmec element into different strains of ST93 MSSA.

Materials and Methods

Bacterial Strains and Identification

Overall 58 ST93 S. aureus were included in the study. The 13 ST93-MSSA included four isolates from remote aboriginal communities in WA, isolated from 1995 to 2003; two isolates from the Northern Territory, isolated in 1992; five isolates from WA, isolated in 2008; and single isolates from Victoria, isolated in 2007, and Queensland, isolated in 2008. The 45 ST93-MRSA included 30 isolates from across Australia from the 2000 to 2008 AGAR Community onset S. aureus programs, and 15 isolates from WA from 2003 to 2009. S. aureus species and methicillin resistance was confirmed by the detection of nuc (thermostable extracellular nuclease) and mecA (methicillin resistance) genes by PCR as previously described [21].

Susceptibility Testing

An antibiogram was performed by disk diffusion on Mueller-Hinton agar according to the Clinical and Laboratory Standards Institute (CLSI) recommendations [22]. A panel of eight antimicrobial drugs was tested: erythromycin (15 µg), tetracycline (30 µg), trimethoprim (5 µg), ciprofloxacin (5 µg), gentamicin (10 µg), rifampin (5 µg), fusidic acid (10 µg), and mupirocin (5 µg). CLSI interpretive criteria [23] were used for all drugs except fusidic acid [24] and mupirocin [25].

PFG

Electrophoresis of chromosomal DNA was performed as previously described [26], using a contour-clamped homogeneous electric field (CHEF) DR III system (Bio-Rad Laboratories Pty Ltd). Chromosomal patterns were examined visually, scanned with a Quantity One device (Bio-Rad Laboratories Pty Ltd), and digitally analyzed using FPQuest (Bio-Rad Laboratories Pty Ltd). S. aureus strain NCTC 8325 was used as a reference strain.

MLST and Spa Typing

Chromosomal DNA for MLST and spa typing was prepared using a DNeasy tissue kit (Qiagen Pty Ltd).

MLST was performed as previously described [27]. The sequences were submitted to http://www.mlst.net/where an allelic profile was generated and an ST assigned.

spa typing, a DNA sequenced-based analysis of the protein A gene variable region was performed as previously described [28] using the nomenclature as described on the Ridom website (http://spa.ridom.de/). Cluster analysis of spa sequences was performed using the spa typing plug-in tool of the BioNumerics software program (version 6.6; Applied Maths, Ghent, Belgium). The analysis compares and aligns sequences via an algorithm based on potential tandem spa repeat duplications, substitutions, and indels (the DSI model) [29]. A minimum spanning tree (MST) was generated from the similarity matrix with the root node assigned to the sequence type with the greatest number of related types. The default software parameters were used for analysis with a bin distance of 1.0%. Thus, the distance between spa types of 99% to 100% similarity was 0, 98% to 99% similarity was assigned a distance of 1, etc., on the MST. For cluster analysis, only spa types separated by an MST distance of ≤1 (i.e., if they were ≥98% similar) were considered closely related and assigned to the same cluster.

DNA Microarray

Arrays and reagents were obtained from Alere Technologies, Jena Germany. The principle of the assay, related procedures, and a list of targets has been described previously [30], [31]. Target genes included species markers, markers for accessory gene regulator (agr) alleles and capsule types, virulence factors, resistance genes, staphylococcal superantigen-like/exotoxin-like genes (set/ssl genes) and genes encoding adhesion proteins and immune evasion factors. Probes for mecA, ugpQ, xylR, kdp, ccr’s, mecI and two probes for mecR were used for SCCmec typing.

SCCmec Typing

The strategy used for SCCmec typing was as previously described [32]. SCCmec nomenclature is used as proposed by the International Working Group on the Classification of Staphylococcal Cassette Chromosome Elements (IWG-SCC) [33]. Briefly, the structural type is indicated by a Roman numeral, with a lowercase letter indicating the subtype, and the ccr complex and the mec complex are indicated by an Arabic numeral and an uppercase letter respectively in parenthesis. Where there is an extra ccr element, this is indicated by “&” and an Arabic numeral designating the ccr type. When there is an extra ccr element present whose precise location is unknown it is indicated by an “&” and ccr number outside the parentheses.

PVL

PCR for the detection of PVL determinants was performed as previously described [34].

PVL Phage Identification

PCRs were performed to detect the six PVL-encoding phages (ΦSa2MW, ΦSa2958, ΦPVL, Φ108PVL, ΦSLT and ΦSA2USA) as previously described [35], [36].

Detection of Allelic Variations in Luks-PV/lukF-PV Genes

Detection of single nucleotide polymorphisms (SNPs) in a defined region of the lukS-PV/lukF-PV genes were performed as previously described [36], [37]. Sequences obtained were compared to the proposed progenitor PVL gene in ΦSLT/ST30.

Quantification of In vitro LukF-PV Expression

PVL is a 2-component exotoxin and both LukS-PV and LukF-PV are required for activity. LukF-PV was measured instead of LukS-PV to obtain an anti-LukF-PV antibody with increased specificity of binding as there was more sequence divergence between lukF-PV and the orthologous 2-component S. aureus exotoxins compared to lukS-PV. To produce recombinant LukF-PV lukF-PV was PCR amplified using primers, forward 5′-CACCATGGCTCAACATATCACAC and reverse 5′-GCTCATAGGATTTTTTTCC. The resulting PCR product was then TOPO cloned into pENTR/SD/D-TOPO (Invitrogen). This plasmid was sequenced using M13 primers to confirm that the insert was present in the correct orientation without mutations. lukF-PV was subsequently cloned using an LR recombination reaction into the expression vector pET-DEST42 (Invitrogen) which introduced a C-terminal 6x-Histidine tag. This expression clone was used to transform Rosetta2 E. coli (Novagen). Soluble recombinant LukF-PV was produced by the Protein Production Unit, Monash University by growth of the expression strain in Auto Induction media at 28°C. The resulting recombinant LukF-PV was purified by Nickel purification followed by gel filtration and eluted in 100 mM NaP04, pH 7.4, 150 mM NaCl buffer. Aliquots were frozen and stored at −80°C. The concentration of recombinant LukF-PV was determined using the 2100 Bioanalyser P230 kit (Agilent).

Quantification of LukF-PV Expression

Bacteria were grown in CCY media (3% yeast extract (Oxoid), 2% Bacto Casamino Acids (Difco), 2.3% sodium pyruvate (Sigma-Aldrich), 0.63% Na2HPO4, 0.041% KH2PO4, pH 6.7). Overnight cultures were diluted 1∶100 into fresh media and then incubated at 37°C with shaking (180 rpm) until stationary phase (OD600 ∼ 1.8). Culture supernatants were harvested by centrifugation and filter sterilized. The LukF-PV expression experiments were performed in at least duplicate for each S. aureus strain. Trichloroacetic acid was added to culture supernatants and incubated at 4°C overnight. Precipitates were then harvested by centrifugation, washed with acetone, air-dried and solubilized in a sample buffer containing 1.7% SDS and 1% 2-mercaptoethanol. The proteins were separated on 12% SDS-PAGE.

A peptide sequence specific to LukF-PV, HWIGNNYKDENRATHT was synthesized and HRP conjugated polyclonal chicken IgY raised against this peptide (Genscript). This antibody was used to detect LukF-PV with enhanced chemiluminescence. Images generated from the western blots were quantitated using GS800 Calibrated Densitometer and Quantity One (BioRad). 50 µg of recombinant LukF-PV was used as an internal standard on each gel, and was the positive control. Results observed with this standard were set to 1.0. All other results were shown as a ratio relative to this standard. RN4220 was used as a negative control.

Dru Typing

Sequence analysis of the mec-associated dru region was performed as previously described [38]. A cluster analysis of dru sequences was performed using the Polymorphic VNTR plug-in tool of the BioNumerics software program (version 6.6; Applied Maths, Ghent, Belgium). The analysis compares and aligns sequences via an algorithm based on potential tandem dru repeat duplications, substitutions, and indels (the DSI model) [29]. A MST was generated from the similarity matrix with the root node assigned to the sequence type with the greatest number of related types. The default software parameters were used for analysis with a bin distance of 1.0%. Thus, the distance between dru types of 99% to 100% similarity was 0, 98% to 99% similarity was assigned a distance of 1, etc., on the MST. For cluster analysis, only dru types separated by an MST distance of ≤1 (i.e., if they were ≥98% similar) were considered closely related and assigned to the same cluster.

Control Strain

The sequenced ST93-IVa [2B] strain JKD6159 (NCBI GenBank Accession No. CP002114 and CP002115) was included in this study for comparison [17].

Results

Susceptibility results, SCCmec typing together with a summary of the resistance genes, spa types (using the Ridom Nomenclature) and dru type are shown in Table 1. Further characterisations are as follows:

Table 1

Characterisation of ST93 isolates.

Region	Reference Number	Year	Specimen	Antibiogram	DNA Microarray Resistance Genotype	PFGE	MLST	spa Sequence	Spa Type	SCCmec type	dru Sequence	Dru Type	lukS/lukF PV
MSSA
NT	WBG 7735	1992	Unknown	ErR	blaZ, blaI, blaR, ermC	B	93	11-17-23-17-17-16-16-16-25	t4178	NA	NA		POSITIVE
NT	WBG 7762	1992	Unknown	ErR	blaZ, blaI, blaR, ermC	A		11-17-16-16-25	t4699	NA	NA		POSITIVE
Qld	UQ40	2008	Unknown		blaZ, blaI, blaR	A		11-17-23-17-17-16-16-25	t202	NA	NA		POSITIVE
Vic	DP 2039	2007	Unknown		blaZ, blaI, blaR	A	93	11-17-23-17-17-16-16-25	t202	NA	NA		POSITIVE
WA	C229T	2003	Throat		blaZ, blaI, blaR	A		11-17-23-17-17-16-16-25-25	t5767	NA	NA		POSITIVE
WA	N126W	2003	Hands	ErR	blaZ, blaI, blaR, ermC	C	93	11-17-23-17-17-16-16-25	t202	NA	NA		POSITIVE
WA	W17S	1995	Skin		blaZ, blaI, blaR	C	93	11-17-23-17-17-16-16-25	t202	NA	NA		POSITIVE
WA	Y113S	1996	Skin	ErR	blaZ, blaI, blaR, ermC	C		11-17-23-17-17-16-16-25	t202	NA	NA		POSITIVE
WA	9506160A	2008	Blood		blaZ, blaI, blaR	A		11-17-23-17-17-16-16-25	t202	NA	NA		POSITIVE
WA	9509712N	2008	Pleural Fl		blaZ, blaI, blaR	A		04-23-17-17-16-16-25	t6485	NA	NA		POSITIVE
WA	9524093R	2008	Buttock		blaZ, blaI, blaR	A		11-17-23-17-17-16-16-25	t202	NA	NA		POSITIVE
WA	9525206A	2008	Blood		blaZ, blaI, blaR	A		11-17-23-17-17-16-16-25	t202	NA	NA		POSITIVE
WA	9529120L	2008	Heel	ErR	blaZ, blaI, blaR, ermC	A		11-17-23-17-17-16-16-25-25	t5767	NA	NA		POSITIVE
MRSA
ACT	SAPTCH92	2000	Blood	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
ACT	SAPTCH53	2008	Boil	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
NSW	SAPRPAH96	2000	Eye	OxR ErR	mecA, blaZ, blaI, blaR, ermC	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-3b	dt3b	POSITIVE
NSW	SAPWH23	2000	Wound	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-3b-4e	dt4d	POSITIVE
NSW	SAPWH39	2000	Wound	OxR ErR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-3b-4e	dt4d	POSITIVE
NSW	SAPWH61	2000	Wound	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
NSW	SAPWH64	2000	Wound	OxR	mecA, blaZ, blaI, blaR, ermC	E	93	11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
NSW	SAPWH94	2000	Wound	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
NSW	SAPWH71	2004	Blood	OxR	mecA, blaZ, blaI, blaR	I	93	11-17-23-17-17-16-16	t6487	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	NEGATIVE
NSW	SAPCRGH95	2007	Blood	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-2a-2g-3b-4e	dt10g	POSITIVE
NSW	SAPRPAH21	2007	Sputum	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
NSW	SAPRPAH7	2008	Ulcer	OxR	mecA, blaZ, blaI, blaR, ermC	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
													POSITIVE
NSW	SAPWH10	2008	Wound	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
													POSITIVE
NSW	SAPWH53	2008	Wound	OxR ErR TmpR	mecA, blaZ, blaI, blaR, msr(A)	K	93	11-17-17-16-16-25	t1811	IVa [2B]	5a-2d-3b	dt3b	NEGATIVE
NT	SAPRDH61	2006	Wound	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
NT	SAPRDH27	2007	Thigh	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
NT	SAPRDH2	2008	Leg	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
Qld	SAPRBH98	2000	Leg	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
Qld	SAPRBH12	2005	Blood	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
Qld	SAPGCH3	2006	Abscess	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
Qld	SAPRBH14	2006	Wound	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
Qld	SAPCBH10	2008	Aspirate	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
Qld	SAPGCH28	2008	Foot	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
Qld	SAPRBH1	2008	Forearm	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
SA	SAPGPSA73	2000	Unknown	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
SA	SAPIMVS24	2006	Abscess	OxR	mecA, blaZ, blaI, blaR	J		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
SA	SAPIMVS31	2008	Boil	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
Vic	SAPRCH74	2000	Wound	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-3b-4e	dt4d	POSITIVE
Vic	SAPAH21	2008	Wound	OxR TeR	mecA, blaZ, blaI, blaR, tetK	D		11-17-23-17-17-16-16-16-25	t4178	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
WA	16790	2003	Axilla	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
WA	16815	2005	Abscess	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
WA	15586	2008	Lip	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-3b	dt3b	POSITIVE
WA	15587	2008	Nose	OxR ErR	mecA, blaZ, blaI, blaR, ermC	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
WA	16414	2008	Thigh	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-3b	dt3b	POSITIVE
WA	16475	2008	ETT	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-4f-3a-2g-3b-4e	dt10i	POSITIVE
WA	17164	2008	Abscess	OxR	mecA, blaZ, blaI, blaR	J		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
WA	18158	2008	Wound	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-3b	dt3b	POSITIVE
WA	18385	2008	Shin	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
WA	18418	2008	Nose	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-3b	dt3b	POSITIVE
WA	20198	2008	Nose	OxR	mecA, blaZ, blaI, blaR	H	93	11-17-23-17-17-16-16-25	t202	V [5C2&5]	5a-2d-4a-0-2d-5b-3a-2g-1c-4e-3e	dt11i	NEGATIVE
WA	SAPRPH48	2008	Wound	OxR	mecA, blaZ, blaI, blaR	J	93	11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
WA	16908	2009	Leg	OxR	mecA, blaZ, blaI, blaR, (dfrA)	F	93	11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
WA	17090	2009	Wound	OxR	mecA, blaZ, blaI, blaR, qacC	G	93	11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
WA	17195	2009	Nose	OxR ErR	mecA, blaZ, blaI, blaR, ermC, qacC	D	93	11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
WA	20548	2009	Abscess	OxR	mecA, blaZ, blaI, blaR	D		11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE
Control Strain
Vic	JKD6159	2004	Blood	OxR	mecA, blaZ, blaI, blaR	D	93	11-17-23-17-17-16-16-25	t202	IVa [2B]	5a-2d-4a-0-2d-5b-3a-2g-3b-4e	dt10a	POSITIVE

Regions: ACT, Australian Capital Territory; NSW, New South Wales; NT, Northern Territory, Qld, Queensland; SA, South Australia; Vic, Victoria; WA, Western Australia.

Antibiogram: Ox, oxacillin; Em, erythromycin; Te, tetracycline.

Resistance Genotype: mecA, alternate penicillin binding protein 2 gene; blaZ, beta lactamase gene; blaI, beta-lactamase repressor (inhibitor) gene; blaR beta-lactamase regulatory protein gene, ermC, erythromycin/clindamycin resistance gene; msr(A), macrolide resistance gene; tetK, tetracycline resistance gene; dfrA, trimethoprim resistance gene; qacC quaternary ammonium compound resistance gene protein C.

PFGE, pulsed field gel electrophoresis; MLST, multilocus sequence type; SCCmec, staphylococcal cassette chromosome mec; spa, Staphylococcus aureus protein gene A; dru, the direct-repeat unit (dru) variable-number tandem repeat region adjacent to IS431 in SCCmec.

lukF/lukS PV, Panton Valentine leucocidin F and S component genes.

Molecular Typing

By PFGE the 58 isolates (13 MSSA and 45 MRSA) had ≥80% similarity with the sequenced JKD6159 strain (Figure 1). Eleven pulsotypes were identified. The MSSA isolates consisted of three pulsotypes, “A” – “C” with 12 of the 13 isolates grouped into two closely related pulsotypes; “A” (9 isolates) and “C” (3 isolates). The MRSA isolates consisted of eight pulsotypes (“C” – “K”) with 39 of the 45 isolates grouped into two closely related pulsotypes; “D” (36 isolates) and “J” (3 isolates). The MSSA pulsotypes “A” and “C” and the MRSA pulsotypes “D” and “J” were 92% related; the difference presumably due to the insertion of the SCCmec type IVa [2B] element into an existing restriction fragment in the two MRSA pulsotypes. Single isolates of closely related MRSA pulsotypes “I” (SAPWH71) and “K” (SAPWH53) lacked a PVL-encoding phage. MRSA pulsotype “H” (20198) also lacked a PVL-encoding phage, however unlike the other MRSA, carried the SCCmec type V element with an additional ccr element [5C2&5]. The remaining MSSA and three MRSA isolates were classified into four unique pulsotypes (pulsotypes B, E, F, G).

Figure 1

Dendrogram of the 58 pulsed-field gel electrophoresis patterns (PFGE) of ST93 (13 MSSA and 45 MRSA).

Sequenced JKD6159 strain was used as the ST93 control. S. aureus strain NCTC 8325 was used as the reference strain.

Isolates representing each pulsotype were identified as ST93 by MLST.

Seven spa types were identified with the majority of isolates characterised as t202 (8/13 MSSA and 42/45 MRSA). The MST algorithm clustered the spa types into two significantly different groups; t202, t4178, t5767; t1811, t4699; plus t6487; and t6485 (Figure 2).

Figure 2

Minimum spanning tree (MST) of the seven ST93 spa types.

Cluster analysis was performed using the spa typing plug-in tool of the BioNumerics program. spa types separated by an MST distance of ≤1 (i.e., if they were ≥98% similar) were considered closely related and assigned to the same cluster. MSSA and MRSA spa types are designated in red and green print respectively. Pulsed-field gel electrophoresis (PFGE) pulsotypes and dru types (dt) are recorded for each spa type.

The β-lactamase operon (blaZ, blaI, blaR) was detected in all isolates (Table 1). Apart from isolate 20198, the MRSA carried mecA as a part of the SCCmec type IVa [2B] element. Carriage of other resistance genes was infrequent and variable. Five of the MSSA were phenotypically erythromycin resistant and carried ermC. Of the five erythromycin resistant MRSA isolates, three carried ermC and one the staphylococcal msr(A) macrolide efflux protein gene. A macrolide resistance gene was not detected in one isolate that demonstrated phenotypic resistance (SAPWH39). Two ermC harbouring MRSA isolates were not phenotypically erythromycin resistant. A single MRSA isolate harboured the tetK tetracycline resistant gene (isolated in Victoria in 2008), and two MRSA isolates carried the quaternary ammonium compound resistance protein C (qacC) gene (isolated in WA in 2009).

The 58 isolates were agr group III and capsule type 8. Although the enterotoxin and tst1 genes were absent from all isolates, the enterotoxin homologue ORF CM14 was present in 34 isolates (4 MSSA and 30 MRSA) (Table S1). All isolates carried the hlb, hld and hlIII hemolysin genes; the staphylokinase (sak), chemotaxis inhibitory protein (chp) and staphylococcal complement inhibitor (scn) genes; and the aur, splA, sspA, sspB, sspP protease genes. Although the gene for a biofilm-associated protein, bap, was absent, the biofilm operon icaACD was present in all isolates. Most isolates carried the leukocidin lukX and lukY genes, the hl, hla hemolysin genes and the splE protease genes.

The staphylococcal superantigen like or exotoxin-like genes (set or ssl genes) and genes encoding MSCRAMMS (microbial surface components recognizing adhesive matrix molecules) and the immune evasion factors were homogeneous and characteristic for ST93 (Tables S2, S3, S4 and S5).

Panton Valentine Leukocidin (PVL)

Apart from three MRSA isolates (SAPWH71, SAPWH53 and 20198) the lukS-PV/lukF-PV genes were detected in all isolates by array hybridisation and PCR. All lukS-PV/lukF-PV positive isolates carried the PVL-encoding phage ΦSa2USA. Using the proposed progenitor PVL gene in ΦSLT/ST30 as a reference sequence, all isolates were similar, having the same R variant SNP profile with three substitutions compared to ΦSLT. This SNP profile is associated with the ΦSa2USA phage.

lukF-PV expression is shown in Figure 3 and was measured to determine if there was a consistent expression profile across different ST93 strains. As expected, ST93 isolates which did not contain lukS-PV/lukF-PV did not express LukF-PV. However, there were three isolates (SAPCRGH95 isolated in NSW, SAPAH21 isolated in Vic, and 15587 isolated in WA) which were PVL positive by array hybridization and PCR but did not express LukF-PV indicating that there may be regulatory differences such as an agr defect in these isolates to account for the absence of LukF-PV. All other isolates produced LukF-PV, with expression levels similar between most strains.

Figure 3

Relative LukF-PV expression in ST93 isolates.

All isolates were tested for LukF-PV expression using western blot and LukF-PV specific antibody. Results are expressed as optical density of test strain relative to a 50 µg control of rLukF-PV that was run on every gel. All experiments were performed with multiple replicates and mean and range is shown. Positive control, rLUKF-PV; negative control, RN4220.

Six dru types were identified (Table 1). The majority of isolates (35/45) were dt10 (33 dt10a, 1 dt10 g and 1 dt10i). The remaining nine SCCmec type IVa [2B] isolates were dru type dt4d (three isolates) and dt3b (6 isolates). The SCCmec type V [5C2&5] isolate (20198) was dt11i. The MST algorithm clustered the six dru types into three significantly different groups; dt10, dt4d plus dt3b, and dt11 (Figure 4). This suggests the SCCmec element may have been acquired on at least three occasions by ST93.

Figure 4

Minimum spanning tree (MST) of the six ST93 dru types.

Cluster analysis was performed using the Polymorphic VNTR plug-in tool of the BioNumerics program. dru types separated by an MST distance of ≤1 (i.e., if they were ≥98% similar) were considered closely related and assigned to the same cluster. Pulsed-field gel electrophoresis (PFGE) pulsotypes and spa types are recorded for each dru type.

Discussion

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 [39]. Although the vertical and horizontal transmission of SCCmec elements into S. aureus has occurred on multiple occasions in the Australian community only a small number of clones have successfully found an ecological niche to predominate over other CA-MRSA clones [40]. PVL-positive ST93-IV [2B] is one such clone, and since 2000 has been reported across Australia and is responsible for the increasing prevalence of CA-MRSA infections nationwide [13].

Conflicting hypotheses have been proposed to explain the molecular evolution of ST93-MRSA. In 2008 Munckhof and colleagues found little genetic diversity within ST93-IV [2B] suggesting it arose from one PVL-positive binary subtype of ST93 MSSA after the acquisition of SCCmec [41]. However in 2009 Tong and colleagues identified multiple spa types in ST93-MRSA and ST93-MSSA and proposed their data supported an early acquisition of SCCmec with subsequent rearrangement of the spa sequence or multiple independent acquisitions of SCCmec and coexistence of MSSA and MRSA versions of the same lineage [42]. Although seven spa types were described in this study, cluster analysis of the seven spa sequences using the Spa typing plug-in tool of the BioNumerics software program shows six of the spa types are closely related and can be assigned to a single cluster (data not shown).

Unlike the Tong study, which examined the spa types of geographically localized ST93 S aureus collected over a short period, the current study examined ST93 S aureus isolated across Australia over sixteen years using a variety of molecular tools, providing greater power to detect unique evolutionary events in geographically diverse regions.

Prior to the isolation of ST93-IV [2B], S aureus surveillance screening of aboriginal people living in 11 remote Western Australian communities identified ST93 as the most prevalent MSSA lineage [20]. Although located in three geographically distant regions of WA, the ST93-MSSA examined from these communities, (W17S isolated in 1995, Y113S in 1996 and C229T and N126W in 2003, and) exhibit limited diversity within their PFGE patterns, spa types and microarray DNA profiles. Their two spa types, t202 [3 isolates (“PFGE C”)] and t5767 (“PFGE A”) are closely related and are assigned to the same cluster. The microarray DNA profiles for the two ST93-MSSA isolated in the Northern Territory in 1992 (WBG7735 and WBG7762) are homogeneous with the four WA remote community ST-93 MSSA. In addition their PFGE patterns are either identical (“PFGE A”) or 90% related (“PFGE B”), and their spa types, t4699 and t4178, are assigned to the same cluster. The DNA microarray profiles for the five ST93-MSSA, (9506160A, 9509712N, 9524093R, 9525206A and 9529120L) isolated in the state’s capital, Perth in 2008 (located 700–2000 km from the remote communities and over 3,000 km from the Northern Territory border) are also homogeneous with the Western Australian remote community strains. The PFGE pattern for these isolates is “PFGE A”. The spa types for four of these strains are t202 (3 isolates) and t5767. The spa type for 9509712N (t6485) cannot be assigned to the same cluster. The PFGE patterns, spa types and microarray DNA profiles for the MSSA isolated on the Australian eastern seaboard (UQ40– Queensland in 2008 and DP2039– Victoria in 2007) are identical to three Perth ST93-MSSA-t202 isolates.

As shown in Figure 1 the MRSA isolates are ≥80% related by PFGE with the majority of isolates falling into pulsotype D. Similar to the MSSA pulsotypes, pulsotype D was dispersed throughout Australia over the eight years. Although rearrangement of the spa sequence has occurred several times, the PFGE patterns and microarray DNA profiles of the 13 ST-93 MSSA isolates suggests the ST93 core and accessory genome is very stable. All carry the PVL-encoding phage ΦSa2USA and their lukS-PV/lukF-PV genes have the same R variant SNP profile. The isolates produce similar expression levels of LukF-PV with no apparent relationship between subtype and PVL expression. The emergence of five different spa types, albeit four types assigned to the same cluster, suggests ST93-MSSA emerged some time ago from a common spa type. As the spa sequences are similar it is not possible to predict the ancestral strain; however one strain, ST93-MSSA-t202, predominates and has successfully disseminated across Australia.

Like ST93-MSSA, ST93-MRSA has multiple spa types; including the closely related t202 and t4178, identified in ST93-MSSA, t1811 and t6487, all of which are assigned to the same cluster. t202 has the largest number of isolates; 42 of the 45 ST93-MRSA. SCCmec and dru typing indicates the SCCmec element has been acquired by ST93-MRSA-t202 on at least three occasions; dt10 (SCCmec type IVa [2B]), dt3b/dt4d (SCCmec type IVa [2B]) and dt11i (SCCmec type V [5C&5]). Unlike ST93-IVa [2B]-t202, ST93-V [5C&5]-t202 does not carry the lukS-PV/lukF-PV genes. The PVL-negative ST93-IVa [2B]-t1811 isolate may have arisen by independent acquisition of SCCmec IVa [2B] or by the subsequent rearrangement of the spa sequence.

As for ST93-MSSA, the PFGE patterns and microarray DNA profiles of the 45 ST-93 MRSA isolates suggests the ST93 core and accessory genome is stable. Forty three of the 45 isolates carry the PVL-encoding phage ΦSa2USA. The lukS-PV/lukF-PV genes have the same R variant SNP profile and produce similar expression levels of LukF-PV as reported in ST93-MSSA.

Apart from the ermC gene which was identified in several early ST93-MSSA and ST93-MRSA isolates, ST93 S. aureus initially carried few antibiotic resistance elements. However since 2008, in addition to mecA and ermC, some isolates of ST93-MRSA have acquired the msr(A) and tetK resistance genes. Although the dfrA gene was not detected by the microarray DNA, SAPWH53 is phenotypically trimethoprim resistant (presumably due to an alternative trimethoprim resistance gene or a different dfrA allele). In addition, the quaternary ammonium compound resistance protein C gene qacC is carried by two isolates. The acquisition of several resistance genes by an epidemic PVL-positive CA-MRSA clone is not unique to ST93-IV [2B]. The USA300 clone (ST8-IV [2B]), initially resistant only to semi-synthetic penicillins and macrolides, is now, frequently resistant to other antimicrobial agents including clindamycin, tetracycline, mupirocin, and the fluoroquinolones; occasionally resistant to gentamicin and trimethoprim-sulfamethoxazole, and may have reduced susceptibility to daptomycin [43].

Single strain outbreaks of ST93-IV [2B] have not been reported in Australian hospitals, however as has been reported in United States hospitals with USA300 [44], ST93-IV [2B] has become a major cause of healthcare-associated/onset infection. In 2008 Munckhof and colleagues reported nearly three quarters of nmMRSA infections in their hospital-based study were healthcare associated, of which ST93-IV [2B] predominated [41].

Conclusion

This study has demonstrated that although multiple rearrangements of the spa sequence have occurred, the core genome in ST93 S. aureus is very stable. Since 2008 PVL-positive ST93-MSSA-t202 has become the predominant ST93-MSSA across Australia. We have shown the emergence of ST93-MRSA has been due to independent acquisitions of different dru-defined type IV and type V SCCmec elements in several spa-defined ST93-MSSA backgrounds. Rearrangement of the spa sequence in ST93-MRSA has subsequently occurred in some of these strains. Although several ST93-MRSA strains have been identified in this study, little genetic diversity was identified for most MRSA isolates, with PVL-positive ST93-IVa [2B]-t202-dt10 predominant across Australia. However to determine if ST93-IVa [2B] t202-dt10 has arisen from one PVL-positive ST93-MSSA-t202, or by independent acquisitions of SCCmec-IVa [2B]-dt10 into multiple PVL-positive ST93-MSSA-t202 strains will require whole genomic sequencing of the isolates. Furthermore, comparative genomic sequencing may further enhance our understanding of the molecular basis for the emergence and increased virulence of ST93 CA-MRSA. At a time when this clone is acquiring additional resistance genes and an increased potential for infections in the healthcare setting, understanding the means for SCCmec acquisition, virulence determinants and transmission dynamics is crucial if we are to prevent this clone from becoming established in hospitals.

Microarray DNA ST93 virulence profile

(DOCX)

Click here for additional data file.

Microarray DNA ST93 ST93 staphylococcal superantigen/enterotoxin-like genes (set/ssl) profile.

(DOCX)

Click here for additional data file.

Microarray DNA ST93 ST93 MSCRAMMs and adhesion profile.

(DOCX)

Click here for additional data file.

Microarray DNA ST93 immunevasion and miscellaneous profile.

(DOCX)

Click here for additional data file.