mecA Locus Diversity in Methicillin-Resistant Staphylococcus aureus Isolates in Brisbane, Australia, and the Development of a Novel Diagnostic Procedure for the Western Samoan Phage Pattern Clone

Abstract

An emerging public health phenomenon is the increasing incidence of methicillin-resistant Staphylococcus aureus (MRSA) infections that are acquired outside of health care facilities. One lineage of community-acquired MRSA (CA-MRSA) is known as the Western Samoan phage pattern (WSPP) clone. The central aim of this study was to develop an efficient genotyping procedure for the identification of WSPP isolates. The approach taken was to make use of the highly variable region downstream of mecA in combination with a single nucleotide polymorphism (SNP) defined by the S. aureus multilocus sequence typing (MLST) database. The premise was that a combinatorial genotyping method that interrogated both a highly variable region and the genomic backbone would deliver a high degree of informative power relative to the number of genetic polymorphisms interrogated. Thirty-five MRSA isolates were used for this study, and their gene contents and order downstream of mecA were determined. The CA-MRSA isolates were found to contain a truncated mecA downstream region consisting of mecA-HVR-IS431 mec-dcs-Ins117, and a PCR-based method for identifying this structure was developed. The hospital-acquired isolates were found to contain eight different mecA downstream regions, three of which were novel. The Minimum SNPs computer software program was used to mine the S. aureus MLST database, and the arcC 272G polymorph was identified as 82% discriminatory for ST-30. A real-time PCR assay was developed to interrogate this SNP. We found that the assay for the truncated mecA downstream region in combination with the interrogation of arcC position 272 provided an unambiguous identification of WSPP isolates.

Community and nosocomial infections caused by methicillin-resistant Staphylococcus aureus (MRSA) are a significant problem worldwide (2, 3, 4, 25). Methicillin resistance is conferred by the mecA gene, which encodes penicillin binding protein 2a (15, 21, 24). The mecA gene is carried by an unusual mobile genetic element termed the staphylococcal-cassette chromosome mec (SCCmec) (12, 13). Enright and coworkers showed that the SCCmec was inserted multiple times in diverse methicillin-susceptible S. aureus (MSSA) clones during the evolution of MRSA (7). Four major SCCmec structural types (I to IV) have been identified based on different ccr and mecA gene complexes (11, 14). Considerable variation between these SCCmec types occurs through the insertion of plasmids and transposable elements downstream of mecA (22).

MRSA strains are classically associated with infections acquired in healthcare facilities. However, recently there have been reports of MRSA infections in the healthy public associated with no or minimal risk factors for infection (1). The majority of these nonmultiresistant community-acquired MRSA (CA-MRSA) strains have been reported globally to carry SCCmec IV (5). The lack of antibiotic selective pressure outside of the hospital environment explains the carriage of SCCmec IV, as no resistance plasmids or transposons have accumulated downstream of mecA (2, 5). These CA-MRSA strains most likely evolved after the acquisition of SCCmec IV by fit community clones of MSSA rather than being descendants of healthcare facility isolates carried into the community (6, 19). Recently, there have been reports of CA-MRSA infections in Southeast Queensland (SEQ), Australia (17, 18). Genotyping indicated that the causative organisms belonged to the Western Samoan phage pattern (WSPP) clone that has previously been associated with community-acquired infections (16). Previously, pulsed-field gel electrophoresis (PFGE) demonstrated that these strains are of pulsotype A (18). Furthermore, this clone appears to have originated in Polynesia and spread to Australia via New Zealand.

We are engaged in the development of efficient typing methods that interrogate polymorphic sites in both highly variable and highly conserved genes and loci. This combined approach is expected to reveal both the long-term evolutionary history of the chromosome and a high-resolution fingerprint resulting from the highly variable region. For this study, the conserved regions chosen were the genes used for multilocus sequence typing (MLST), while the variable regions used were the mecA downstream regions. Our hypothesis was that interrogating the highly variable region downstream of mecA and a single nucleotide polymorphism (SNP) defined by computerized mining of the MLST database would enable the reliable determination of whether an isolate belongs to the WSPP clone. Therefore, the aims of this study were to determine the diversity of the mecA downstream regions in a collection of SEQ hospital-acquired and CA-MRSA strains, to find an SNP discriminatory for the WSPP clone by using the S. aureus MLST database, and to determine the combinatorial power of interrogating the SNP and the mecA downstream region.

MATERIALS AND METHODS

MRSA isolates.

The 35 MRSA isolates used for this study are listed in Table 1. They are all from SEQ, Australia, and were selected from a collection that was previously characterized by PFGE and by their gene contents downstream of mecA (10, 18). One isolate conforms to the description of UK EMRSA-1, -4, or -11 by its MLST sequence type (7).

TABLE 1.

MRSA isolates used for this study

Isolate no.	Isolate identification	Origin of acquisition	PFGE type
1	A803355	Community	A0
2	66460/98	Community	A0
3	D828570	Community	A0
4	E822547	Community	A0
5	F810539	Community	A0
6	B826559	Community	A0
7	A806533	Community	A0
8	E803534	Community	A0
9	D817541	Community	A0
10	A830538	Community	A0
11	I825560	Community	A0
12	A823547	Community	A1
13	C810534	Community	A1
14	D821552	Community	A2
15	E802537	Community	A3
16	68284/98	Community	A5
17	E822485	Hospital	B
18	J710566	Nursing home	C
19	F829549	Community	D
20	C801535	Hospital	D
21	D828534	Hospital	E
22	B827549	Nursing home	E
23	PA01M18489	Hospital	EMRSA-1, -4, -11a
24	K704540	Local	F
25	K705613	Hospital	F2
26	K711532	Hospital	F3
27	K714372	Hospital	F4
28	K703484	Hospital	G1
29	I823541	Hospital	G2
30	E804531	Hospital	I
31	E812560	Hospital	J
32	B8-31	Pathology center	K
33	IPOOM14235	Hospital	O
34	IP01M2046	Hospital	P1
35	IP01M1081	Hospital	Q

^aThis isolate was essentially identical to that of EMRSA-1, -4, or -11 described by Enright et al. (7).

Cultivation and DNA isolation.

All isolates were stored in glycerol stock solutions at −80°C. They were routinely grown overnight at 37°C on brain heart infusion agar plates (Oxoid, Hampshire, England). For the isolation of genomic DNA, a single colony was selected, suspended in 5 ml of brain heart infusion broth (Oxoid), and cultured overnight with aeration at 37°C. The chromosomal DNA was extracted from 1 ml of this suspension by use of a Qiagen DNA extraction kit (Qiagen, Victoria, Australia) according to the manufacturer's instructions, with lysostaphin (Sigma Chemical Company, St. Louis, Mo.) at 200 μg/ml used for the lysis step. The eluted DNA solutions were frozen at −20°C for subsequent use.

Primer design.

The sequences for elements associated with the mecA downstream region were obtained through the GenBank database web site (http://www.ncbi.nlm.nih.gov). Their accession numbers are as follows: for pT181 and pI258, AB037671; and for pUB110, IS431, the downstream common sequence (dcs), and Ins117, AF181950. Primers (Table 2) were designed to amplify across the junctions of these mobile elements which were previously shown to be associated with the mecA downstream region. The sizes of the respective amplicons for each primer combination used are shown in Table 3, and the locations of the primers in the elements are shown in Fig. 1.

TABLE 2.

Primers used for mapping the mecA downstream region

GenBank accession no.	Target	Primer	Orientation	Sequence (5′-3′)	Source or reference
AF181950	mecA	mecA P1	Forward	ATC GAT GGT AAA GGT TGG C	22
AF181950	HVR	HVR P1	Forward	ATG TCC CAA GCT CCA TTT TG	22
		HVR P2	Reverse	TGG AGC TTG GGA CAT AAA TG	22
AF181950	IS431	IS P1	Forward	AAG GGA ATC TTC TGT ATG AAC	22
		IS P4	Forward	CAG GTC TCT TCA GAT CTA CG	22
		IS P3	Reverse	TTA CTT TAG CCA TTG CTA CC	22
J01764	pT181	pT181 F2	Forward	GCA ATC AAT GAA CCA AGA CAG C	This study
		D F1	Reverse	CAC GAG ATG AAA TGA TTT	10
L29436	pI258	D R2	Forward	TTT ATA CGT AAA CCA GTC GG	10
		pI258 F	Reverse	GAC TCA TTG TTG CTT CCT GG	This study
AF181950	pUB110	pUB110 F1	Forward	TTG ATG ACA CAG AAG AAG GC	This study
		pUB110 R1	Reverse	CTC ATT CCC TTT TCA GAT AA	This study
AF181950	Ins117	Ins117 R2	Reverse	GTT TTT TCA GCC GCT T	This study
M18086	IS256	D F3	Eithera	ACT AAT GGA AAA TCA ACG	10
		D R3	Eithera	TTT TTT TCT GAT AAT AAA CG	10
AF181950	dcs	DCS F1	Forward	AGA CTG TGG ACA AAC TGA TT	This study
		MDV F1	Forward	GCT TGG GTA ACT TAT CAT GG	22
		MDV R5	Reverse	CAT GGC TAT GAT TTA GTA GC	22
		Ins117 R1	Reverse	CTA AAT ATA GTA AAT TAC GG	10

^aThe insertion sequence occurs in both orientations.

TABLE 3.

Amplicon sizes based on GenBank sequences AB037671, AB033763, and AF181950

Amplicon	Forward primer	Reverse primer	Amplicon size (bp)
mecA-HVR	mecA P1	HVR P2	2,226
mecA-Ins117	mecA P1	Ins117 R2	6,000
HVR-IS431	HVR P1	IS P3	1,572
IS431-pT181	IS P4	D F1	744
pT181-IS431	pT181 F2	IS P3	918
IS431-pI258	IS P4	pI258 F	223
pI258-IS431	D R2	IS P3	1,837
pT181-pI258	pT181 F2	pI258 F	3,157
HVR-pI258a	HVR P1	pI258 F	1,969
HVR-pT181a	HVR P1	D F1	4,800
HVR-pUB110a	HVR P1	pUB110 R1	2,008
IS431-pUB110	IS P4	pUB110 R1	261
pUB110-IS431	pUB110 F1	IS P3	647
IS431-dcs	IS P4	MDV R5	1,059

^aNo submitted sequence data were available for the full-length fragment, so distances were calculated by the addition of known lengths.

FIG. 1.

Binding sites for primers used to determine mecA downstream arrangements. The scale is in kilobases.

DNA amplification.

PCR amplifications were performed on an MJ Research thermocycler (GeneWorks, Adelaide, Australia) in 0.2-ml PCR tubes containing 20 mM Tris-HCl, 100 mM KCl, 1 mM dithiothreitol, 0.1 mM EDTA, 0.5% (vol/vol) Tween, 2.25 mM MgCl₂, a 0.2 mM concentration of each deoxynucleoside triphosphate (PCR nucleotide mix; Roche Diagnostics, Castle Hill, Australia), 0.5 μl each of the forward and reverse primers, 0.7 U of polymerase enzyme mix (Roche Expand Long Template PCR system; Roche Diagnostics), and 5 μl of 20-ng/μl purified DNA template solution in a 50-μl total volume. The amplifications were carried out at the following temperature profile: 94°C for 4 min; 30 cycles of 94°C for 30 s, 50°C for 30 s, and 72°C for 2 min 30 s; and a final extension step of 72°C for 10 min. For longer amplicons (over 5 kb), the following temperature profile was used: 94°C for 4 min; 10 cycles of 94°C for 30 s, 50°C for 30 s, and 68°C for 5 min; 20 cycles of 94°C for 30 s, 50°C for 30 s, and 68°C for 5 min plus 20 s/cycle; and a final extension step of 72°C for 10 min.

Analysis of PCR products.

PCR products were visualized in a 1.0% agarose gel electrophoresed in TBE buffer (90 mM Tris-borate, 2 mM EDTA) at 110 V for 30 to 40 min in the presence of ethidium bromide. PCR products were sized against a molecular weight marker (Marker X; Roche Diagnostics). Eight microliters of product was adequate to determine the presence and quality of the PCR products.

Sequence type determination.

MLST was performed as specified by Enright et al. (6). In brief, PCR products were purified by use of a QIAquick PCR purification kit (Qiagen) and quantitated in a 1% agarose gel by comparison to a DNA mass ladder (Marker VIII; Roche Diagnostics). Purified products were used in sequencing reactions with both forward and reverse primers. Each sequencing mix contained an estimated 100 ng of purified PCR product, 3.2 pmol of primer, 4 μl of ABI Big Dye terminator mix (Applied Biosystems, Victoria, Australia), and water to a final volume of 20 μl. Samples were electrophoresed at the Australian Genome Research Facility at the University of Queensland, St. Lucia, Australia. The sequences obtained were compared with the sequences at the MLST web site (http://www.mlst.net/) to assign sequence types (STs).

Identification of informative SNPs.

The computer program Minimum SNPs (23) was used to analyze the S. aureus MLST database in order to find an SNP that was discriminatory for the WSPP clone. At the time this work was carried out, the following STs were available for download: 1 to 103, 109, 120, 121, 123, 124, 134, 145, 169, 178, 182, 188 to 190, 192 to 205, 207 to 215, 217, 220 to 222, 225, 228, 231, 238 to 241, 243, 246 to 247, and 250. These were used for all analyses.

Kinetic PCR.

Kinetic PCR is an allele-specific PCR performed in a real-time format (8). It involves performing two or more reactions, each of which contains an allele-specific primer and a common primer. Each allele-specific primer is a perfect match with one of the alleles defined by the states of the SNP. The reaction in which there is a perfect match reaches the threshold level of amplification product in fewer cycles than the reaction in which there is a primer mismatch, i.e., the cycles to threshold (C_T) value for the matched reaction is smaller. The sign of difference in C_T values (ΔC_T) between the reactions indicates the base present at the SNP. Product detection was done with Sybr Green. The primers used are listed in Table 4. These were designed with Primer Express software, version 2.0 (Applied Biosystems), with the mismatched nucleotide at the 3′end of the forward primer and one common reverse primer. The primers were designed to amplify an 84-bp product. The reactions were performed in an ABI Prism 7000 sequence detection system (Applied Biosystems). Reaction volumes were 25 μl and included a 0.2 μM concentration (each) of forward and reverse primer (E@sy Oligo; Genset Oligos, Lismore, Australia), 2.5 mM MgCl₂ (Roche Diagnostics), 0.2 mM PCR nucleotide mix (Roche Diagnostics), 0.5 U of Taq DNA polymerase (Roche Diagnostics), 1× PCR buffer (100 mM Tris, 500 mM KCl, pH 8.3) (Roche Diagnostics), 0.125 μl of SYBR Green (1:100 working solution) (Molecular Probes, Quantum Scientific, Queensland, Australia), and 5 to 10 ng of genomic DNA. The cycling conditions were 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 56°C for 1 min. All reactions were done in duplicate.

TABLE 4.

Primers used for kinetic PCR

Primer	Orientation	Sequence (5′-3′)
RT1	Forward	GAAGAATTACAAAAAGAACAGCCAGG
RT2	Forward	GAAGAATTACAAAAAGAACAGCCAGA
RT3	Reverse	GGTAGTGGTGACGCAACTACTTTTCTA

RESULTS

mecA downstream regions in healthcare-related isolates.

An objective of this study was to determine the diversity of the mecA downstream regions in a collection of SEQ healthcare-related MRSA isolates. Huygens and coworkers had previously shown that these isolates contain a variety of antibiotic resistance plasmids and insertion sequences (10). In order to build on those findings, we determined the arrangements of the genes and other elements by using PCR to assay for the presence of junctions between these elements.

The MRSA isolate ATCC 49476 was previously shown to contain a large complement of mobile elements downstream of mecA (10, 11). The elements detected were plasmid pT181, which carries the gene for resistance to tetracycline, plasmid pI258, which carries the gene for resistance to mercury, and the insertion sequence IS256. These antibiotic resistance plasmids have been shown to be flanked by copies of IS431 (9, 22). Therefore, this isolate offered many potential junctions between mobile elements to be amplified that would validate the PCR primers. The downstream region of mecA in this isolate was determined through the amplification of a series of junctions between the conserved and mobile elements present. The fragments amplified were as predicted from the known gene order (11) and are shown in Fig. 2 as arrangement J.

FIG. 2.

Downstream mecA arrangement of S. aureus ATCC 49476 (arrangement J). The scale is shown in kilobases.

In the group of healthcare-related isolates, five isolates (isolate numbers 17, 28, 29, 31, and 33) contained the pUB110 plasmid. The junctions that were amplified to determine the arrangement of the elements were mecA-HVR, HVR-IS431 mec, HVR-pUB110, IS431 mec-pUB110, pUB110-IS431, and pUB110-dcs. The amplicon size for the HVR-pUB110 junction was 1.9 kb, which indicates that the plasmid had inserted directly downstream of IS431 mec. The IS431-Ins117 junction was negative for four of the five isolates. Therefore, the downstream mecA arrangement for four isolates containing pUB110 was mecA-HVR-IS431 mec-pUB110-IS431-dcs, with the remaining isolate containing the IS431-Ins117 junction (Fig. 3, arrangements A and B).

FIG. 3.

Downstream mecA arrangements of hospital-acquired isolates. The scale is shown in kilobases.

Two isolates (numbers 21 and 22) contained the pT181 plasmid. The following junctions were amplified to define only one arrangement: mecA-HVR, HVR-IS431 mec, HVR-pT181, dcs-pT181, IS431-pT181, and pT181-IS431. These isolates produced an amplicon of approximately 4.8 kb between the HVR and pT181. This amplicon was composed of 1.85 kb from the forward primer site in the HVR to the end of IS431 mec, a portion of the dcs, and 1.29 kb from the start of the left-hand side flanking IS431 of pT181 to the reverse primer site within pT181. The remaining distance making up the 4.8 kb was approximately 1.6 kb of the dcs. The mecA downstream regions described by Oliveira et al. (22) support the theory that pT181 has inserted 1.6 kb into the dcs, as they show the insertion of an IS431-flanked pI258 1.6 kb into the dcs. This pT181 insertion was further clarified by the amplification of a fragment of 2.4 kb between the forward primer (dcs F1) of the dcs and pT181. Therefore, the confirmed order of elements in pT181-containing isolates is mecA-HVR-IS431 mec and 1.6 kb of dcs-IS431-pT181-IS431 (Fig. 3, arrangement C).

There were four isolates (isolate numbers 24, 25, 26, and 27) that were shown to carry the pI258 plasmid. Only one arrangement was determined by amplification of the following junctions: mecA-HVR, HVR-IS431 mec, HVR-pI258, IS431-pI258, pI258-IS431, and pI258-dcs. The HVR-pI258 amplicon was 1.9 kb, which indicated that the plasmid had inserted directly downstream of IS431 mec. There was no indication from mec-associated mobile element typing that these isolates contained Ins117, and therefore we did not screen for the insertion sequence. The final arrangement for these isolates was determined to be mecA-HVR-IS431 mec-pI258-IS431-dcs (Fig. 3, arrangement D).

Arrangements E, F, G, and H (Fig. 3) were found to be similar to arrangement I, which was typically found in the community-acquired isolates (5). However, these arrangements could be differentiated from I mainly due to either the absence of the dcs region (arrangement H), the absence of Ins117 (arrangement F), or a variation in the length of the dcs region (arrangements E and G). These arrangements were found in hospital-acquired isolate numbers 18, 20, 23, and 34.

Specific primer set for the mecA downstream arrangement found in CA-MRSA.

CA-MRSA isolates have previously been reported to carry the type IV SSCmec (5), and a defining feature of this is a truncated form of the mecA downstream region (shown as arrangement I in Fig. 3). A previous analysis of the gene contents of these isolates (10) suggested that the CA-MRSA isolates in this collection also have this mecA downstream structure.

An important aim of this study was to develop a set of primers that would rapidly discriminate community-acquired from hospital-acquired isolates. This test was designed to allow the differentiation of community-acquired isolates from hospital-acquired isolates with similar phenotypes. This was achieved by designing a minimal number of primers separated by reasonable distances covering the downstream mecA region of these isolates (Fig. 4A). This primer set was designed to generate a specific PCR product pattern for the community-acquired isolates or for other isolates that contained an identical mobile element arrangement. The first primer pair amplifies a product common to all known MRSA strains and serves as a positive control for mecA with the HVR immediately distal. The primers used were mecA P1 and HVR P2, which produce a fragment of 2.2 kb. The second primer pair, HVR P1 and MDV R5, amplifies a fragment of 2.8 kb from the HVR across the IS431 mec and 0.94 kb into the dcs. This differentiates the shorter mecA downstream region characteristic of CA-MRSA by demonstrating (i) the presence of a full-length HVR and (ii) that no other plasmids have been inserted between the IS431 mec and the dcs. The final primer pair, primers IS P4 and Ins117 R2, was designed to amplify from the IS431 mec to Ins117. The amplicon produced is 2.3 kb, and again, if the arrangement of this region is different from that of the community-acquired arrangement, then the fragment either will not amplify or will be considerably larger. All of the CA-MRSA isolates were subjected to this specific primer set and all produced the three fragments of the expected sizes. Examples are shown in Fig. 4B. This confirms that the mecA downstream arrangement associated with community-acquired isolates was present. Three hospital-acquired isolates, namely 30, 32, and 35, also produced the three amplicons of the expected sizes, indicating a mecA downstream region identical to that associated with the community-acquired isolates.

FIG. 4.

(A) Positions of the three primer pairs used to discriminate the downstream mecA arrangements associated with community-acquired isolates. The scale is shown in kilobases. (B) Amplification products from the primer set specific for arrangement I. Lanes 1 to 3, isolate 1; lanes 4 to 6, isolate 7; lanes 7 to 9, isolate 13; lanes 10 to 12, isolate 31. The isolates shown in lanes 1 to 9 were community acquired and show the amplification of the three fragments at the expected sizes. Lanes 10 to 12 show results for a hospital-acquired isolate that does not have arrangement I: a truncated dcs is present (arrangement B). MW, molecular weight marker (Roche X).

MLST of representative MRSA isolates.

MLST was carried out with five representative community-acquired WSPP MRSA isolates as well as four hospital-acquired MRSA isolates. Table 5 lists the results of MLST of the MRSA isolates tested. The WSPP lineage was previously found to belong to PFGE pulsotype A (10, 18). These results are consistent with previous findings that WSPP belongs to ST-30 and that other isolates belong to different STs (19).

TABLE 5.

MLST of representative isolates

Isolate no.	Origin of acquisition	PFGE pulsotype	ST
19	Community	D	88
20	Hospital	D	SLV of 88c
7	Community	A0	30
1	Community	A0	30
2	Community	A0	30
23	Hospital	EMRSA-1, -4, -11a	239
Typing isolateb	Community	A6	30

^aThe pulsotype of this isolate was identical to that of EMRSA-1, -4, or -11, as described by Enright et al. (5).

^bThis isolate was not used for the mapping of the downstream mec elements but only for MLST.

^cSLV, single locus variant.

Kinetic PCRs of CA-MRSA isolates.

One of the aims of this study was to develop a kinetic PCR assay that distinguishes WSPP MRSA clones from other MRSA clones. With the aid of the Minimum SNPs software, we identified arcC 272G as an informative polymorph. This SNP provides 82% discrimination, i.e., 82% of known STs do not have a G at this SNP. The STs apart from ST-30 that have a G at arcC position 272 are ST-2, ST-17, ST-19, ST-24, ST-30, ST-31, ST-32, ST-33, ST-36, ST-37, ST-38, ST-39, ST-40, ST-41, ST-43, ST-57, ST-74, ST-77, ST-86, ST-196, ST-200, ST-210, ST-238, ST-239, ST-240, ST-241, ST-243, and ST-246.

The results of the kinetic PCRs performed with all of the MRSA isolates are listed in Table 6. The kinetic PCRs were validated by using two isolates with known STs, namely isolate number 2, of ST-30, and isolate number 19, of ST-88. The G polymorph is present at arcC position 272 for isolate number 2, and the A polymorph is present at this position for isolate number 19. After optimization of the experimental conditions, the ΔC_T values were always in the correct orientation for both isolates. We found that, as expected, all of the pulsotype A isolates possessed a G at arcC position 272.

TABLE 6.

Kinetic PCR results

Isolate no.	PFGE pulsotype	Origin of acquisition	ST-30-specific primer CT	Non-ST-30-specific primer CT	ΔCT	mecA arrangement
1	A0	Community	15.9	20.7	4.8	I
2	A0	Community	17.4	20.2	2.8	I
3	A0	Community	16.6	21.8	5.2	I
4	A0	Community	19.9	26.2	6.3	I
5	A0	Community	16.8	21.8	5.0	I
6	A0	Community	16.3	21.8	5.5	I
7	A0	Community	18.0	25.2	7.2	I
8	A0	Community	16.7	21.5	4.8	I
9	A0	Community	19.5	24.7	5.2	I
10	A0	Community	17.3	22.4	5.1	I
11	A0	Community	16.6	20.8	4.2	I
12	A1	Community	18.7	25.0	6.3	I
13	A1	Community	18.9	27.2	8.3	I
14	A2	Community	17.8	26.4	8.6	I
15	A3	Community	18.4	27.5	9.1	I
16	A5	Community	19.0	25.2	6.2	I
17	B	Hospital	20.9	27.4	6.5b	B
18	C	Nursing home	18.1	16.4	−1.7	G
19	D	Community	21.2	17.6	−3.6	I
20	D	Hospital	23.1	21.7	−1.4	H
21	E	Hospital	22.1	21.4	−0.7	C
22	E	Nursing home	21.6	19.0	−2.6	C
23	EMRSA-1, -4, -11a	Hospital	16.8	24.2	7.4b	F
24	F	Local	20.8	27.4	6.6b	D
25	F2	Hospital	17.6	22.2	4.6b	D
26	F3	Hospital	20.00	28.0	8b	D
27	F4	Hospital	16.7	23.4	6.7b	D
28	G1	Hospital	19.6	28.3	8.7b	A
29	G2	Hospital	19.2	31.1	11.9b	A
30	I	Hospital	22.2	17.5	−4.7	I
31	J	Hospital	19.3	26.9	7.6b	A
32	K	Pathology center	27.1	17.9	−9.2	I
33	O	Hospital	16.5	20.7	4.2b	A
34	P1	Hospital	16.4	15.5	−0.9	E
35	Q	Hospital	17.5	15.4	−2.1	I

^aSee the work of Enright et al. for an explanation (6).

^bSimilar ΔC_T as the pulsotype A isolates, but with different mecA arrangement.

Informative power of arcC 272G polymorph in combination with mecA downstream region characteristic of CA-MRSA.

All CA-MRSA isolates had the downstream mecA arrangement of mecA-HVR-IS431 mec-dcs-Ins117. Three healthcare-related isolates of diverse pulsotypes (isolates 30, 32, and 35) were also shown to have the same downstream mecA arrangement. However, kinetic PCRs showed that these isolates had a C rather than an A at arcC position 272. Furthermore, 10 non-pulsotype A isolates (isolates 17, 23 to 29, 31, and 33) were found to have the arcC 272G polymorph, which was also found in the pulsotype A ST-30 (WSPP) isolates. However, none of these possessed the mecA downstream region arrangement found in CA-MRSA. Therefore, the combination of the detection of the arcC 272G polymorph and the downstream mecA arrangement of mecA-HVR-IS431 mec-dcs-Ins117 is 100% specific and 100% sensitive for pulsotype A ST-30 CA-MRSA with this collection of isolates.

DISCUSSION

The rapid accumulation of comparative genomic information is facilitating the development of targeted and efficient microbial genotyping procedures that are based on known genetic polymorphisms. A combinatorial strategy of interrogating polymorphisms in highly variable regions and also in the more stable genomic backbone has the potential to provide highly informative epidemiological fingerprints. We have applied this strategy to the development of a type-specific diagnostic procedure for the WSPP MRSA clone that is associated with community-acquired infections.

One major aspect of this work was the further elucidation of the diversity of the mecA downstream region in MRSA. This represents a continuation of analyses reported by Huygens et al. (10). The hospital isolates used in this study were previously shown to be diverse (10, 18). In our collection, 23 isolates were from nosocomial infections primarily acquired from postoperative infections in 21 patients. There were eight different arrangements found for these isolates, and the results broadly reflect the gene orders found by Oliveira et al. (22). Three arrangements have not previously been reported (arrangements C, E, and G), with the variation being primarily in the dcs region. Notably, in one of these (arrangement E), there was a 1.6-kb deletion in the dcs. Another interesting arrangement was arrangement C, which contained the pT181 plasmid. The amplification of the junctions of these isolates revealed that the plasmid had inserted 1.6 kb of sequence into the dcs. A similar insertion event was reported by Oliveira et al. (22), but with the pI258 plasmid inserted instead. This further elucidation of the diversity in this region greatly assisted us in designing a specific procedure for discriminating the WSPP clone from other MRSA clones.

Seventeen of the 35 isolates used in this study were classified as community acquired. The majority were from patients with soft-tissue abscesses. SNP interrogation and full sequence type determination confirmed that the WSPP clone is of ST-30, as reported by Okuma and coworkers (19). The finding that each isolate had the downstream mecA arrangement of mecA-HVR-IS431 mec-dcs-Ins117 is consistent with studies of CA-MRSA globally. This arrangement is equivalent to the 20-kb type IV SCCmec which is nearly always found in CA-MRSA (5). A possible reason for this is that S. aureus strains in the community are not subjected to the intense selective pressures of diverse antibiotics that are seen in healthcare-related facilities. It is likely that the WSPP clone emerged as a result of a recent introduction of SCCmec type IV into a successful ST-30 MSSA clone. While it may not be surprising that the closely related WSPP clone members all share the same arrangement of genes downstream of mecA, the two pulsotype D isolates (isolates 19 and 20) are of considerable interest because isolate 19 is the only non-pulsotype A CA-MRSA strain in this collection and appears to possess exactly the same mecA downstream region as the WSPP isolates, while the hospital-acquired isolate 20 has lost the dcs primer binding site.

The proliferation of large comparative gene sequence databases, such as those used for MLST, provides a valuable resource for the development of rationally designed typing methodologies. Members of our research group developed a computer program (Minimum SNPs) that can mine the MLST database and identify highly informative SNP sets (23). The software is able to take entire MLST databases as input and provide as output sets of SNPs that either discriminate a user-defined ST or provide a high Simpson's index of diversity with respect to the MLST database. For this study, the Minimum SNPs software was used to identify a single SNP that discriminates MRSA members of the ST-30 clonal complex from other MRSA isolates. Currently, the MLST web site includes typing data for a wide range of bacterial pathogens. Minimum SNPs provides a generalized, straightforward, and effective means of identifying SNPs in multilocus comparative sequence databases that may be interrogated in combination with highly variable regions to provide highly informative genotypes.

Diversity in the SCCmec elements is extensive and complex (14). A classification scheme for SCCmec elements is becoming accepted. This is based upon diversity both in the immediate vicinity of mecA and in the ccr gene cassette. Oliveira and de Lencastre (20) and Okuma and coworkers (19) have described generalized PCR-based methods for typing the SCCmec elements. It may be true that the primer pairs specific for the type IV SCCmec would provide a similar performance in combination with the arcC 272G polymorph interrogation identifying the WSPP clone. However, in the case of the typing method described by Oliveira and de Lencastre (20), the SCCmec type IV-specific reaction relies on the absence of a product from the reaction that targets the pls region, and a negative PCR may not be appropriate for inclusion in an assay for a particular clone. Also, it is clear that extensive variability exists within the SCCmec types, and this remains to be fully understood. Our work has revealed more of the nature of the diversity of the mecA downstream regions of the SCCmec elements. In particular, the presence of the Ins117 element has been shown to differentiate the mecA downstream region in the WSPP clone from very similar mecA downstream regions in other MRSA lineages. The use of generalized methods for typing the SCCmec elements in combination with the interrogation of SNPs defined by MLST databases may in the future prove to be a very effective generalized method for MRSA typing.

In the collection of isolates used for this study, only members of the pulsotype A WSPP clone possessed a G at arcC position 272 and arrangement I downstream of mecA. This reflects the apparent evolutionary history of the WSPP clone; it is an ST-30 strain that has acquired this characteristic version of the type IV SSCmec and also the ability to cause community-acquired infections. We do not know if this 100% specificity would be retained with a larger and more diverse collection of isolates. However, for an isolate to give a false positive it would need to possess arrangement I downstream of mecA and also to be one of the small minority of STs that have a G at arcC position 272. We have not yet found any such isolates in SEQ.

In conclusion, we have analyzed the mecA downstream regions of a variety of SEQ isolates and used this information for the design of a specific assay for the WSPP CA-MRSA clone. This work has demonstrated a novel and generalizable approach to the design of very specific DNA-based diagnostic procedures.